diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6693 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 9500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0010526315789473684,
+      "grad_norm": 5.129565589745107,
+      "learning_rate": 4.5e-06,
+      "loss": 3.831922149658203,
+      "step": 10
+    },
+    {
+      "epoch": 0.002105263157894737,
+      "grad_norm": 1.8060645220873177,
+      "learning_rate": 9.5e-06,
+      "loss": 3.500311279296875,
+      "step": 20
+    },
+    {
+      "epoch": 0.003157894736842105,
+      "grad_norm": 1.5274659739720347,
+      "learning_rate": 1e-05,
+      "loss": 3.072599983215332,
+      "step": 30
+    },
+    {
+      "epoch": 0.004210526315789474,
+      "grad_norm": 1.439529323295704,
+      "learning_rate": 1e-05,
+      "loss": 2.6167324066162108,
+      "step": 40
+    },
+    {
+      "epoch": 0.005263157894736842,
+      "grad_norm": 0.6226358274800787,
+      "learning_rate": 1e-05,
+      "loss": 2.2310859680175783,
+      "step": 50
+    },
+    {
+      "epoch": 0.00631578947368421,
+      "grad_norm": 0.579778608698981,
+      "learning_rate": 1e-05,
+      "loss": 2.0481679916381834,
+      "step": 60
+    },
+    {
+      "epoch": 0.007368421052631579,
+      "grad_norm": 0.5255768872176885,
+      "learning_rate": 1e-05,
+      "loss": 1.9059135437011718,
+      "step": 70
+    },
+    {
+      "epoch": 0.008421052631578947,
+      "grad_norm": 0.5641137907790387,
+      "learning_rate": 1e-05,
+      "loss": 1.7455703735351562,
+      "step": 80
+    },
+    {
+      "epoch": 0.009473684210526316,
+      "grad_norm": 0.6643724228276126,
+      "learning_rate": 1e-05,
+      "loss": 1.7015619277954102,
+      "step": 90
+    },
+    {
+      "epoch": 0.010526315789473684,
+      "grad_norm": 0.5713809514426843,
+      "learning_rate": 1e-05,
+      "loss": 1.5777016639709474,
+      "step": 100
+    },
+    {
+      "epoch": 0.011578947368421053,
+      "grad_norm": 0.5597405118979538,
+      "learning_rate": 1e-05,
+      "loss": 1.5415353775024414,
+      "step": 110
+    },
+    {
+      "epoch": 0.01263157894736842,
+      "grad_norm": 0.4758333039729566,
+      "learning_rate": 1e-05,
+      "loss": 1.493780517578125,
+      "step": 120
+    },
+    {
+      "epoch": 0.01368421052631579,
+      "grad_norm": 0.45630310192212664,
+      "learning_rate": 1e-05,
+      "loss": 1.4095694541931152,
+      "step": 130
+    },
+    {
+      "epoch": 0.014736842105263158,
+      "grad_norm": 0.9108908227389748,
+      "learning_rate": 1e-05,
+      "loss": 1.4157570838928222,
+      "step": 140
+    },
+    {
+      "epoch": 0.015789473684210527,
+      "grad_norm": 0.4673730229456233,
+      "learning_rate": 1e-05,
+      "loss": 1.350106430053711,
+      "step": 150
+    },
+    {
+      "epoch": 0.016842105263157894,
+      "grad_norm": 0.5223769302039626,
+      "learning_rate": 1e-05,
+      "loss": 1.1846101760864258,
+      "step": 160
+    },
+    {
+      "epoch": 0.017894736842105262,
+      "grad_norm": 0.48745189601377703,
+      "learning_rate": 1e-05,
+      "loss": 1.2197025299072266,
+      "step": 170
+    },
+    {
+      "epoch": 0.018947368421052633,
+      "grad_norm": 0.43040595966792417,
+      "learning_rate": 1e-05,
+      "loss": 1.2024085998535157,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3698951775181469,
+      "learning_rate": 1e-05,
+      "loss": 1.1386910438537599,
+      "step": 190
+    },
+    {
+      "epoch": 0.021052631578947368,
+      "grad_norm": 0.4411702491308663,
+      "learning_rate": 1e-05,
+      "loss": 1.1117079734802247,
+      "step": 200
+    },
+    {
+      "epoch": 0.022105263157894735,
+      "grad_norm": 0.3708081724031713,
+      "learning_rate": 1e-05,
+      "loss": 1.088566780090332,
+      "step": 210
+    },
+    {
+      "epoch": 0.023157894736842106,
+      "grad_norm": 0.4406720948389181,
+      "learning_rate": 1e-05,
+      "loss": 1.0874387741088867,
+      "step": 220
+    },
+    {
+      "epoch": 0.024210526315789474,
+      "grad_norm": 0.3873880830276917,
+      "learning_rate": 1e-05,
+      "loss": 1.1115880966186524,
+      "step": 230
+    },
+    {
+      "epoch": 0.02526315789473684,
+      "grad_norm": 0.38806667496267544,
+      "learning_rate": 1e-05,
+      "loss": 1.1190993309020996,
+      "step": 240
+    },
+    {
+      "epoch": 0.02631578947368421,
+      "grad_norm": 0.3989310609683637,
+      "learning_rate": 1e-05,
+      "loss": 1.053086280822754,
+      "step": 250
+    },
+    {
+      "epoch": 0.02736842105263158,
+      "grad_norm": 0.3640974504512374,
+      "learning_rate": 1e-05,
+      "loss": 1.0668390274047852,
+      "step": 260
+    },
+    {
+      "epoch": 0.028421052631578948,
+      "grad_norm": 0.3785285911712781,
+      "learning_rate": 1e-05,
+      "loss": 1.081777000427246,
+      "step": 270
+    },
+    {
+      "epoch": 0.029473684210526315,
+      "grad_norm": 0.3645950395756694,
+      "learning_rate": 1e-05,
+      "loss": 1.1065789222717286,
+      "step": 280
+    },
+    {
+      "epoch": 0.030526315789473683,
+      "grad_norm": 0.36223820924716266,
+      "learning_rate": 1e-05,
+      "loss": 1.0594990730285645,
+      "step": 290
+    },
+    {
+      "epoch": 0.031578947368421054,
+      "grad_norm": 0.34754318158030956,
+      "learning_rate": 1e-05,
+      "loss": 1.009918785095215,
+      "step": 300
+    },
+    {
+      "epoch": 0.03263157894736842,
+      "grad_norm": 0.39831625777745766,
+      "learning_rate": 1e-05,
+      "loss": 1.0630088806152345,
+      "step": 310
+    },
+    {
+      "epoch": 0.03368421052631579,
+      "grad_norm": 0.5492823887352639,
+      "learning_rate": 1e-05,
+      "loss": 1.0556842803955078,
+      "step": 320
+    },
+    {
+      "epoch": 0.034736842105263156,
+      "grad_norm": 0.39653533605000163,
+      "learning_rate": 1e-05,
+      "loss": 1.0781284332275392,
+      "step": 330
+    },
+    {
+      "epoch": 0.035789473684210524,
+      "grad_norm": 0.36509939099563055,
+      "learning_rate": 1e-05,
+      "loss": 1.048421859741211,
+      "step": 340
+    },
+    {
+      "epoch": 0.03684210526315789,
+      "grad_norm": 0.42109561658578115,
+      "learning_rate": 1e-05,
+      "loss": 1.0543712615966796,
+      "step": 350
+    },
+    {
+      "epoch": 0.037894736842105266,
+      "grad_norm": 0.38877588128194784,
+      "learning_rate": 1e-05,
+      "loss": 1.0506539344787598,
+      "step": 360
+    },
+    {
+      "epoch": 0.03894736842105263,
+      "grad_norm": 0.36244453874240107,
+      "learning_rate": 1e-05,
+      "loss": 1.0279829978942872,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3899162246283317,
+      "learning_rate": 1e-05,
+      "loss": 1.0612271308898926,
+      "step": 380
+    },
+    {
+      "epoch": 0.04105263157894737,
+      "grad_norm": 0.792780755469385,
+      "learning_rate": 1e-05,
+      "loss": 1.0633869171142578,
+      "step": 390
+    },
+    {
+      "epoch": 0.042105263157894736,
+      "grad_norm": 0.5284715949266973,
+      "learning_rate": 1e-05,
+      "loss": 1.0811123847961426,
+      "step": 400
+    },
+    {
+      "epoch": 0.0431578947368421,
+      "grad_norm": 0.36690524387473483,
+      "learning_rate": 1e-05,
+      "loss": 1.029583740234375,
+      "step": 410
+    },
+    {
+      "epoch": 0.04421052631578947,
+      "grad_norm": 0.3405692085001304,
+      "learning_rate": 1e-05,
+      "loss": 1.0555681228637694,
+      "step": 420
+    },
+    {
+      "epoch": 0.045263157894736845,
+      "grad_norm": 0.3606451850180366,
+      "learning_rate": 1e-05,
+      "loss": 1.0335403442382813,
+      "step": 430
+    },
+    {
+      "epoch": 0.04631578947368421,
+      "grad_norm": 0.3952670426627856,
+      "learning_rate": 1e-05,
+      "loss": 1.0426573753356934,
+      "step": 440
+    },
+    {
+      "epoch": 0.04736842105263158,
+      "grad_norm": 0.36727132253562994,
+      "learning_rate": 1e-05,
+      "loss": 1.0272632598876954,
+      "step": 450
+    },
+    {
+      "epoch": 0.04842105263157895,
+      "grad_norm": 0.38043192529364367,
+      "learning_rate": 1e-05,
+      "loss": 1.0343192100524903,
+      "step": 460
+    },
+    {
+      "epoch": 0.049473684210526316,
+      "grad_norm": 0.3553258558273771,
+      "learning_rate": 1e-05,
+      "loss": 1.0337108612060546,
+      "step": 470
+    },
+    {
+      "epoch": 0.05052631578947368,
+      "grad_norm": 0.372236793032702,
+      "learning_rate": 1e-05,
+      "loss": 1.0002114295959472,
+      "step": 480
+    },
+    {
+      "epoch": 0.05157894736842105,
+      "grad_norm": 0.35668459763117616,
+      "learning_rate": 1e-05,
+      "loss": 1.0200517654418946,
+      "step": 490
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 0.3490627557461056,
+      "learning_rate": 1e-05,
+      "loss": 1.0097460746765137,
+      "step": 500
+    },
+    {
+      "epoch": 0.05368421052631579,
+      "grad_norm": 0.3142385034895546,
+      "learning_rate": 1e-05,
+      "loss": 1.0037628173828126,
+      "step": 510
+    },
+    {
+      "epoch": 0.05473684210526316,
+      "grad_norm": 0.3619214862173559,
+      "learning_rate": 1e-05,
+      "loss": 1.0973596572875977,
+      "step": 520
+    },
+    {
+      "epoch": 0.05578947368421053,
+      "grad_norm": 0.3902312672919626,
+      "learning_rate": 1e-05,
+      "loss": 1.0703372955322266,
+      "step": 530
+    },
+    {
+      "epoch": 0.056842105263157895,
+      "grad_norm": 0.3827645858944265,
+      "learning_rate": 1e-05,
+      "loss": 1.119423007965088,
+      "step": 540
+    },
+    {
+      "epoch": 0.05789473684210526,
+      "grad_norm": 0.3903006853246503,
+      "learning_rate": 1e-05,
+      "loss": 1.1221393585205077,
+      "step": 550
+    },
+    {
+      "epoch": 0.05894736842105263,
+      "grad_norm": 0.3888788557636633,
+      "learning_rate": 1e-05,
+      "loss": 1.1268860816955566,
+      "step": 560
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37691058586137594,
+      "learning_rate": 1e-05,
+      "loss": 1.1172764778137207,
+      "step": 570
+    },
+    {
+      "epoch": 0.061052631578947365,
+      "grad_norm": 0.3962363794127834,
+      "learning_rate": 1e-05,
+      "loss": 1.148810577392578,
+      "step": 580
+    },
+    {
+      "epoch": 0.06210526315789474,
+      "grad_norm": 0.36003705654899615,
+      "learning_rate": 1e-05,
+      "loss": 1.1295830726623535,
+      "step": 590
+    },
+    {
+      "epoch": 0.06315789473684211,
+      "grad_norm": 0.41036268725648356,
+      "learning_rate": 1e-05,
+      "loss": 1.1855300903320312,
+      "step": 600
+    },
+    {
+      "epoch": 0.06421052631578947,
+      "grad_norm": 0.4353462759352186,
+      "learning_rate": 1e-05,
+      "loss": 1.2502850532531737,
+      "step": 610
+    },
+    {
+      "epoch": 0.06526315789473684,
+      "grad_norm": 0.7027107228957341,
+      "learning_rate": 1e-05,
+      "loss": 1.241795539855957,
+      "step": 620
+    },
+    {
+      "epoch": 0.06631578947368422,
+      "grad_norm": 0.38406883373511647,
+      "learning_rate": 1e-05,
+      "loss": 1.2793304443359375,
+      "step": 630
+    },
+    {
+      "epoch": 0.06736842105263158,
+      "grad_norm": 0.6688008709362958,
+      "learning_rate": 1e-05,
+      "loss": 1.2975357055664063,
+      "step": 640
+    },
+    {
+      "epoch": 0.06842105263157895,
+      "grad_norm": 0.4870402021626517,
+      "learning_rate": 1e-05,
+      "loss": 1.2863228797912598,
+      "step": 650
+    },
+    {
+      "epoch": 0.06947368421052631,
+      "grad_norm": 0.4789365037156072,
+      "learning_rate": 1e-05,
+      "loss": 1.3121292114257812,
+      "step": 660
+    },
+    {
+      "epoch": 0.07052631578947369,
+      "grad_norm": 0.4249289667365913,
+      "learning_rate": 1e-05,
+      "loss": 1.3259175300598145,
+      "step": 670
+    },
+    {
+      "epoch": 0.07157894736842105,
+      "grad_norm": 0.65676501689554,
+      "learning_rate": 1e-05,
+      "loss": 1.3344863891601562,
+      "step": 680
+    },
+    {
+      "epoch": 0.07263157894736842,
+      "grad_norm": 0.4282154338817059,
+      "learning_rate": 1e-05,
+      "loss": 1.3299467086791992,
+      "step": 690
+    },
+    {
+      "epoch": 0.07368421052631578,
+      "grad_norm": 0.46223182471205226,
+      "learning_rate": 1e-05,
+      "loss": 1.366246509552002,
+      "step": 700
+    },
+    {
+      "epoch": 0.07473684210526316,
+      "grad_norm": 0.39766756701321343,
+      "learning_rate": 1e-05,
+      "loss": 1.3451871871948242,
+      "step": 710
+    },
+    {
+      "epoch": 0.07578947368421053,
+      "grad_norm": 0.40900253954134147,
+      "learning_rate": 1e-05,
+      "loss": 1.3738602638244628,
+      "step": 720
+    },
+    {
+      "epoch": 0.07684210526315789,
+      "grad_norm": 0.43925417914640763,
+      "learning_rate": 1e-05,
+      "loss": 1.383627700805664,
+      "step": 730
+    },
+    {
+      "epoch": 0.07789473684210527,
+      "grad_norm": 0.46019955787293465,
+      "learning_rate": 1e-05,
+      "loss": 1.3509666442871093,
+      "step": 740
+    },
+    {
+      "epoch": 0.07894736842105263,
+      "grad_norm": 0.44872485307857357,
+      "learning_rate": 1e-05,
+      "loss": 1.3666427612304688,
+      "step": 750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4168661453527556,
+      "learning_rate": 1e-05,
+      "loss": 1.3329523086547852,
+      "step": 760
+    },
+    {
+      "epoch": 0.08105263157894736,
+      "grad_norm": 0.39483526379862016,
+      "learning_rate": 1e-05,
+      "loss": 1.3404739379882813,
+      "step": 770
+    },
+    {
+      "epoch": 0.08210526315789474,
+      "grad_norm": 0.4027045862960815,
+      "learning_rate": 1e-05,
+      "loss": 1.3771724700927734,
+      "step": 780
+    },
+    {
+      "epoch": 0.08315789473684211,
+      "grad_norm": 0.4250328488579122,
+      "learning_rate": 1e-05,
+      "loss": 1.3159744262695312,
+      "step": 790
+    },
+    {
+      "epoch": 0.08421052631578947,
+      "grad_norm": 0.3863025736069049,
+      "learning_rate": 1e-05,
+      "loss": 1.3153966903686523,
+      "step": 800
+    },
+    {
+      "epoch": 0.08526315789473685,
+      "grad_norm": 0.3794385382688299,
+      "learning_rate": 1e-05,
+      "loss": 1.3049957275390625,
+      "step": 810
+    },
+    {
+      "epoch": 0.0863157894736842,
+      "grad_norm": 0.3923592960136275,
+      "learning_rate": 1e-05,
+      "loss": 1.2895668029785157,
+      "step": 820
+    },
+    {
+      "epoch": 0.08736842105263158,
+      "grad_norm": 0.3972373746266566,
+      "learning_rate": 1e-05,
+      "loss": 1.2735277175903321,
+      "step": 830
+    },
+    {
+      "epoch": 0.08842105263157894,
+      "grad_norm": 0.38703288927278834,
+      "learning_rate": 1e-05,
+      "loss": 1.2897537231445313,
+      "step": 840
+    },
+    {
+      "epoch": 0.08947368421052632,
+      "grad_norm": 0.3847485591417034,
+      "learning_rate": 1e-05,
+      "loss": 1.2797724723815918,
+      "step": 850
+    },
+    {
+      "epoch": 0.09052631578947369,
+      "grad_norm": 0.411875160914232,
+      "learning_rate": 1e-05,
+      "loss": 1.262526035308838,
+      "step": 860
+    },
+    {
+      "epoch": 0.09157894736842105,
+      "grad_norm": 0.42983503070040374,
+      "learning_rate": 1e-05,
+      "loss": 1.2575672149658204,
+      "step": 870
+    },
+    {
+      "epoch": 0.09263157894736843,
+      "grad_norm": 0.4143550833492262,
+      "learning_rate": 1e-05,
+      "loss": 1.2420232772827149,
+      "step": 880
+    },
+    {
+      "epoch": 0.09368421052631579,
+      "grad_norm": 0.4094647869883051,
+      "learning_rate": 1e-05,
+      "loss": 1.2448989868164062,
+      "step": 890
+    },
+    {
+      "epoch": 0.09473684210526316,
+      "grad_norm": 0.39643601027185765,
+      "learning_rate": 1e-05,
+      "loss": 1.2363014221191406,
+      "step": 900
+    },
+    {
+      "epoch": 0.09578947368421052,
+      "grad_norm": 0.40125922761930166,
+      "learning_rate": 1e-05,
+      "loss": 1.2585180282592774,
+      "step": 910
+    },
+    {
+      "epoch": 0.0968421052631579,
+      "grad_norm": 0.3889609835337118,
+      "learning_rate": 1e-05,
+      "loss": 1.2770251274108886,
+      "step": 920
+    },
+    {
+      "epoch": 0.09789473684210526,
+      "grad_norm": 0.3878257510334844,
+      "learning_rate": 1e-05,
+      "loss": 1.2342063903808593,
+      "step": 930
+    },
+    {
+      "epoch": 0.09894736842105263,
+      "grad_norm": 0.40383266926577155,
+      "learning_rate": 1e-05,
+      "loss": 1.242508602142334,
+      "step": 940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3971115120579083,
+      "learning_rate": 1e-05,
+      "loss": 1.2583754539489747,
+      "step": 950
+    },
+    {
+      "epoch": 0.10105263157894737,
+      "grad_norm": 0.3539562726239424,
+      "learning_rate": 1e-05,
+      "loss": 1.2484332084655763,
+      "step": 960
+    },
+    {
+      "epoch": 0.10210526315789474,
+      "grad_norm": 0.41227605004446527,
+      "learning_rate": 1e-05,
+      "loss": 1.2285819053649902,
+      "step": 970
+    },
+    {
+      "epoch": 0.1031578947368421,
+      "grad_norm": 0.47456847590786266,
+      "learning_rate": 1e-05,
+      "loss": 1.2464988708496094,
+      "step": 980
+    },
+    {
+      "epoch": 0.10421052631578948,
+      "grad_norm": 0.36810144667048633,
+      "learning_rate": 1e-05,
+      "loss": 1.2326807975769043,
+      "step": 990
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 0.3894473717694802,
+      "learning_rate": 1e-05,
+      "loss": 1.230353546142578,
+      "step": 1000
+    },
+    {
+      "epoch": 0.10631578947368421,
+      "grad_norm": 0.39241910502330524,
+      "learning_rate": 1e-05,
+      "loss": 1.2199285507202149,
+      "step": 1010
+    },
+    {
+      "epoch": 0.10736842105263159,
+      "grad_norm": 0.3974911457522689,
+      "learning_rate": 1e-05,
+      "loss": 1.2336506843566895,
+      "step": 1020
+    },
+    {
+      "epoch": 0.10842105263157895,
+      "grad_norm": 0.455364431016719,
+      "learning_rate": 1e-05,
+      "loss": 1.2420743942260741,
+      "step": 1030
+    },
+    {
+      "epoch": 0.10947368421052632,
+      "grad_norm": 0.4345817306363754,
+      "learning_rate": 1e-05,
+      "loss": 1.2048468589782715,
+      "step": 1040
+    },
+    {
+      "epoch": 0.11052631578947368,
+      "grad_norm": 0.3856529201471913,
+      "learning_rate": 1e-05,
+      "loss": 1.2200773239135743,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11157894736842106,
+      "grad_norm": 0.35788417168939973,
+      "learning_rate": 1e-05,
+      "loss": 1.2127140045166016,
+      "step": 1060
+    },
+    {
+      "epoch": 0.11263157894736842,
+      "grad_norm": 0.3848585578501117,
+      "learning_rate": 1e-05,
+      "loss": 1.2093496322631836,
+      "step": 1070
+    },
+    {
+      "epoch": 0.11368421052631579,
+      "grad_norm": 0.38062519869646566,
+      "learning_rate": 1e-05,
+      "loss": 1.2119295120239257,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11473684210526315,
+      "grad_norm": 0.37454599624618173,
+      "learning_rate": 1e-05,
+      "loss": 1.20859375,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11578947368421053,
+      "grad_norm": 0.4436679866388131,
+      "learning_rate": 1e-05,
+      "loss": 1.2168142318725585,
+      "step": 1100
+    },
+    {
+      "epoch": 0.1168421052631579,
+      "grad_norm": 0.4010256367494713,
+      "learning_rate": 1e-05,
+      "loss": 1.2059317588806153,
+      "step": 1110
+    },
+    {
+      "epoch": 0.11789473684210526,
+      "grad_norm": 0.37715595465600615,
+      "learning_rate": 1e-05,
+      "loss": 1.2100018501281737,
+      "step": 1120
+    },
+    {
+      "epoch": 0.11894736842105263,
+      "grad_norm": 0.3722674452607559,
+      "learning_rate": 1e-05,
+      "loss": 1.173952007293701,
+      "step": 1130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3572349214226328,
+      "learning_rate": 1e-05,
+      "loss": 1.1537309646606446,
+      "step": 1140
+    },
+    {
+      "epoch": 0.12105263157894737,
+      "grad_norm": 0.32466803707265374,
+      "learning_rate": 1e-05,
+      "loss": 1.1372849464416503,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12210526315789473,
+      "grad_norm": 0.368130829469978,
+      "learning_rate": 1e-05,
+      "loss": 1.1409140586853028,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1231578947368421,
+      "grad_norm": 0.35626305226063965,
+      "learning_rate": 1e-05,
+      "loss": 1.11646728515625,
+      "step": 1170
+    },
+    {
+      "epoch": 0.12421052631578948,
+      "grad_norm": 0.34180196520782097,
+      "learning_rate": 1e-05,
+      "loss": 1.071176528930664,
+      "step": 1180
+    },
+    {
+      "epoch": 0.12526315789473685,
+      "grad_norm": 0.33065026528413527,
+      "learning_rate": 1e-05,
+      "loss": 1.0419276237487793,
+      "step": 1190
+    },
+    {
+      "epoch": 0.12631578947368421,
+      "grad_norm": 0.3413557539933989,
+      "learning_rate": 1e-05,
+      "loss": 1.0239316940307617,
+      "step": 1200
+    },
+    {
+      "epoch": 0.12736842105263158,
+      "grad_norm": 0.31896029240841567,
+      "learning_rate": 1e-05,
+      "loss": 0.9914163589477539,
+      "step": 1210
+    },
+    {
+      "epoch": 0.12842105263157894,
+      "grad_norm": 0.3234597344172216,
+      "learning_rate": 1e-05,
+      "loss": 0.9772817611694335,
+      "step": 1220
+    },
+    {
+      "epoch": 0.12947368421052632,
+      "grad_norm": 0.32713607429000163,
+      "learning_rate": 1e-05,
+      "loss": 0.9316439628601074,
+      "step": 1230
+    },
+    {
+      "epoch": 0.13052631578947368,
+      "grad_norm": 0.307817077958078,
+      "learning_rate": 1e-05,
+      "loss": 0.9130254745483398,
+      "step": 1240
+    },
+    {
+      "epoch": 0.13157894736842105,
+      "grad_norm": 0.2930690259584948,
+      "learning_rate": 1e-05,
+      "loss": 0.9260563850402832,
+      "step": 1250
+    },
+    {
+      "epoch": 0.13263157894736843,
+      "grad_norm": 0.28211266658635614,
+      "learning_rate": 1e-05,
+      "loss": 0.8800353050231934,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1336842105263158,
+      "grad_norm": 0.3180735021763057,
+      "learning_rate": 1e-05,
+      "loss": 0.8895734786987305,
+      "step": 1270
+    },
+    {
+      "epoch": 0.13473684210526315,
+      "grad_norm": 0.302750450306546,
+      "learning_rate": 1e-05,
+      "loss": 0.8559271812438964,
+      "step": 1280
+    },
+    {
+      "epoch": 0.13578947368421052,
+      "grad_norm": 0.2991432619758276,
+      "learning_rate": 1e-05,
+      "loss": 0.8617020606994629,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1368421052631579,
+      "grad_norm": 0.3392010475402778,
+      "learning_rate": 1e-05,
+      "loss": 0.8292947769165039,
+      "step": 1300
+    },
+    {
+      "epoch": 0.13789473684210526,
+      "grad_norm": 0.2896126425438961,
+      "learning_rate": 1e-05,
+      "loss": 0.82211275100708,
+      "step": 1310
+    },
+    {
+      "epoch": 0.13894736842105262,
+      "grad_norm": 0.28710492304121205,
+      "learning_rate": 1e-05,
+      "loss": 0.8157825469970703,
+      "step": 1320
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2871905927145198,
+      "learning_rate": 1e-05,
+      "loss": 0.7862643241882324,
+      "step": 1330
+    },
+    {
+      "epoch": 0.14105263157894737,
+      "grad_norm": 0.2821122063508978,
+      "learning_rate": 1e-05,
+      "loss": 0.8182851791381835,
+      "step": 1340
+    },
+    {
+      "epoch": 0.14210526315789473,
+      "grad_norm": 0.3069348630427839,
+      "learning_rate": 1e-05,
+      "loss": 0.7791574954986572,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1431578947368421,
+      "grad_norm": 0.2826513953939523,
+      "learning_rate": 1e-05,
+      "loss": 0.7499025821685791,
+      "step": 1360
+    },
+    {
+      "epoch": 0.14421052631578948,
+      "grad_norm": 0.3640726274503572,
+      "learning_rate": 1e-05,
+      "loss": 0.8512861251831054,
+      "step": 1370
+    },
+    {
+      "epoch": 0.14526315789473684,
+      "grad_norm": 0.28706582671563924,
+      "learning_rate": 1e-05,
+      "loss": 0.8040716171264648,
+      "step": 1380
+    },
+    {
+      "epoch": 0.1463157894736842,
+      "grad_norm": 0.2887007748240138,
+      "learning_rate": 1e-05,
+      "loss": 0.7845241069793701,
+      "step": 1390
+    },
+    {
+      "epoch": 0.14736842105263157,
+      "grad_norm": 0.2857885839056229,
+      "learning_rate": 1e-05,
+      "loss": 0.7895279884338379,
+      "step": 1400
+    },
+    {
+      "epoch": 0.14842105263157895,
+      "grad_norm": 0.28115201742861623,
+      "learning_rate": 1e-05,
+      "loss": 0.8022497177124024,
+      "step": 1410
+    },
+    {
+      "epoch": 0.14947368421052631,
+      "grad_norm": 0.36141069309573115,
+      "learning_rate": 1e-05,
+      "loss": 0.7775122642517089,
+      "step": 1420
+    },
+    {
+      "epoch": 0.15052631578947367,
+      "grad_norm": 0.28046748369731184,
+      "learning_rate": 1e-05,
+      "loss": 0.771450138092041,
+      "step": 1430
+    },
+    {
+      "epoch": 0.15157894736842106,
+      "grad_norm": 0.42813359693492475,
+      "learning_rate": 1e-05,
+      "loss": 0.84803466796875,
+      "step": 1440
+    },
+    {
+      "epoch": 0.15263157894736842,
+      "grad_norm": 0.3307722958354092,
+      "learning_rate": 1e-05,
+      "loss": 0.9220462799072265,
+      "step": 1450
+    },
+    {
+      "epoch": 0.15368421052631578,
+      "grad_norm": 0.3209023952391098,
+      "learning_rate": 1e-05,
+      "loss": 0.9727088928222656,
+      "step": 1460
+    },
+    {
+      "epoch": 0.15473684210526314,
+      "grad_norm": 0.3355481706974916,
+      "learning_rate": 1e-05,
+      "loss": 0.9762910842895508,
+      "step": 1470
+    },
+    {
+      "epoch": 0.15578947368421053,
+      "grad_norm": 0.321777119672059,
+      "learning_rate": 1e-05,
+      "loss": 0.997312068939209,
+      "step": 1480
+    },
+    {
+      "epoch": 0.1568421052631579,
+      "grad_norm": 0.376820135360389,
+      "learning_rate": 1e-05,
+      "loss": 1.0065011978149414,
+      "step": 1490
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.3248443532885038,
+      "learning_rate": 1e-05,
+      "loss": 1.0136358261108398,
+      "step": 1500
+    },
+    {
+      "epoch": 0.15894736842105264,
+      "grad_norm": 0.353481320904289,
+      "learning_rate": 1e-05,
+      "loss": 1.040408992767334,
+      "step": 1510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38059582977924694,
+      "learning_rate": 1e-05,
+      "loss": 1.0027363777160645,
+      "step": 1520
+    },
+    {
+      "epoch": 0.16105263157894736,
+      "grad_norm": 0.35015076552913627,
+      "learning_rate": 1e-05,
+      "loss": 1.0386922836303711,
+      "step": 1530
+    },
+    {
+      "epoch": 0.16210526315789472,
+      "grad_norm": 0.3228048466722668,
+      "learning_rate": 1e-05,
+      "loss": 1.036502742767334,
+      "step": 1540
+    },
+    {
+      "epoch": 0.1631578947368421,
+      "grad_norm": 0.3270304730309021,
+      "learning_rate": 1e-05,
+      "loss": 1.0355409622192382,
+      "step": 1550
+    },
+    {
+      "epoch": 0.16421052631578947,
+      "grad_norm": 0.3285486278859695,
+      "learning_rate": 1e-05,
+      "loss": 1.0135831832885742,
+      "step": 1560
+    },
+    {
+      "epoch": 0.16526315789473683,
+      "grad_norm": 0.5133953646112632,
+      "learning_rate": 1e-05,
+      "loss": 1.0526947975158691,
+      "step": 1570
+    },
+    {
+      "epoch": 0.16631578947368422,
+      "grad_norm": 0.3414584783614906,
+      "learning_rate": 1e-05,
+      "loss": 1.0366554260253906,
+      "step": 1580
+    },
+    {
+      "epoch": 0.16736842105263158,
+      "grad_norm": 0.3213301337507541,
+      "learning_rate": 1e-05,
+      "loss": 1.0404515266418457,
+      "step": 1590
+    },
+    {
+      "epoch": 0.16842105263157894,
+      "grad_norm": 0.32844883930307683,
+      "learning_rate": 1e-05,
+      "loss": 1.0619461059570312,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1694736842105263,
+      "grad_norm": 0.3633960653841322,
+      "learning_rate": 1e-05,
+      "loss": 0.9977890014648437,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1705263157894737,
+      "grad_norm": 0.32763723702059705,
+      "learning_rate": 1e-05,
+      "loss": 1.053870964050293,
+      "step": 1620
+    },
+    {
+      "epoch": 0.17157894736842105,
+      "grad_norm": 0.3529921729977246,
+      "learning_rate": 1e-05,
+      "loss": 1.0153541564941406,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1726315789473684,
+      "grad_norm": 0.34716488140776836,
+      "learning_rate": 1e-05,
+      "loss": 1.034360122680664,
+      "step": 1640
+    },
+    {
+      "epoch": 0.1736842105263158,
+      "grad_norm": 0.3584988617028858,
+      "learning_rate": 1e-05,
+      "loss": 1.0256132125854491,
+      "step": 1650
+    },
+    {
+      "epoch": 0.17473684210526316,
+      "grad_norm": 0.335968191456543,
+      "learning_rate": 1e-05,
+      "loss": 1.0312114715576173,
+      "step": 1660
+    },
+    {
+      "epoch": 0.17578947368421052,
+      "grad_norm": 0.3130726732719835,
+      "learning_rate": 1e-05,
+      "loss": 1.0257963180541991,
+      "step": 1670
+    },
+    {
+      "epoch": 0.17684210526315788,
+      "grad_norm": 0.3149633737400575,
+      "learning_rate": 1e-05,
+      "loss": 1.0275050163269044,
+      "step": 1680
+    },
+    {
+      "epoch": 0.17789473684210527,
+      "grad_norm": 0.33561768553144145,
+      "learning_rate": 1e-05,
+      "loss": 1.0279806137084961,
+      "step": 1690
+    },
+    {
+      "epoch": 0.17894736842105263,
+      "grad_norm": 0.3076217072018895,
+      "learning_rate": 1e-05,
+      "loss": 1.028935146331787,
+      "step": 1700
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.35380664789534094,
+      "learning_rate": 1e-05,
+      "loss": 1.0285554885864259,
+      "step": 1710
+    },
+    {
+      "epoch": 0.18105263157894738,
+      "grad_norm": 0.3200387028937822,
+      "learning_rate": 1e-05,
+      "loss": 1.0226529121398926,
+      "step": 1720
+    },
+    {
+      "epoch": 0.18210526315789474,
+      "grad_norm": 0.31825042083818195,
+      "learning_rate": 1e-05,
+      "loss": 1.032950210571289,
+      "step": 1730
+    },
+    {
+      "epoch": 0.1831578947368421,
+      "grad_norm": 0.31239956122549917,
+      "learning_rate": 1e-05,
+      "loss": 1.0238349914550782,
+      "step": 1740
+    },
+    {
+      "epoch": 0.18421052631578946,
+      "grad_norm": 0.3198787668647505,
+      "learning_rate": 1e-05,
+      "loss": 1.058379554748535,
+      "step": 1750
+    },
+    {
+      "epoch": 0.18526315789473685,
+      "grad_norm": 0.32644757771125277,
+      "learning_rate": 1e-05,
+      "loss": 1.0524209976196288,
+      "step": 1760
+    },
+    {
+      "epoch": 0.1863157894736842,
+      "grad_norm": 0.389036234008021,
+      "learning_rate": 1e-05,
+      "loss": 1.0207571029663085,
+      "step": 1770
+    },
+    {
+      "epoch": 0.18736842105263157,
+      "grad_norm": 0.3258748231424236,
+      "learning_rate": 1e-05,
+      "loss": 1.0353762626647949,
+      "step": 1780
+    },
+    {
+      "epoch": 0.18842105263157893,
+      "grad_norm": 0.31542831524196,
+      "learning_rate": 1e-05,
+      "loss": 1.0521798133850098,
+      "step": 1790
+    },
+    {
+      "epoch": 0.18947368421052632,
+      "grad_norm": 0.3138383766525853,
+      "learning_rate": 1e-05,
+      "loss": 1.0374223709106445,
+      "step": 1800
+    },
+    {
+      "epoch": 0.19052631578947368,
+      "grad_norm": 0.31211031018273094,
+      "learning_rate": 1e-05,
+      "loss": 1.0329813957214355,
+      "step": 1810
+    },
+    {
+      "epoch": 0.19157894736842104,
+      "grad_norm": 0.29295687070458576,
+      "learning_rate": 1e-05,
+      "loss": 1.027329158782959,
+      "step": 1820
+    },
+    {
+      "epoch": 0.19263157894736843,
+      "grad_norm": 0.3253322210625686,
+      "learning_rate": 1e-05,
+      "loss": 1.033319091796875,
+      "step": 1830
+    },
+    {
+      "epoch": 0.1936842105263158,
+      "grad_norm": 0.4056170386267799,
+      "learning_rate": 1e-05,
+      "loss": 1.0567522048950195,
+      "step": 1840
+    },
+    {
+      "epoch": 0.19473684210526315,
+      "grad_norm": 0.3085723413269434,
+      "learning_rate": 1e-05,
+      "loss": 1.0215108871459961,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1957894736842105,
+      "grad_norm": 0.37270482500218244,
+      "learning_rate": 1e-05,
+      "loss": 1.029244327545166,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1968421052631579,
+      "grad_norm": 0.31734657001974687,
+      "learning_rate": 1e-05,
+      "loss": 1.0313403129577636,
+      "step": 1870
+    },
+    {
+      "epoch": 0.19789473684210526,
+      "grad_norm": 0.32487222822048256,
+      "learning_rate": 1e-05,
+      "loss": 1.0314602851867676,
+      "step": 1880
+    },
+    {
+      "epoch": 0.19894736842105262,
+      "grad_norm": 0.3591269426129111,
+      "learning_rate": 1e-05,
+      "loss": 1.0066585540771484,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36757190710932325,
+      "learning_rate": 1e-05,
+      "loss": 1.0303743362426758,
+      "step": 1900
+    },
+    {
+      "epoch": 0.20105263157894737,
+      "grad_norm": 0.3238980551162356,
+      "learning_rate": 1e-05,
+      "loss": 1.0154070854187012,
+      "step": 1910
+    },
+    {
+      "epoch": 0.20210526315789473,
+      "grad_norm": 0.31389649725944707,
+      "learning_rate": 1e-05,
+      "loss": 1.0318874359130858,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2031578947368421,
+      "grad_norm": 0.3165937543347964,
+      "learning_rate": 1e-05,
+      "loss": 1.0315516471862793,
+      "step": 1930
+    },
+    {
+      "epoch": 0.20421052631578948,
+      "grad_norm": 0.32643802559768836,
+      "learning_rate": 1e-05,
+      "loss": 1.0183154106140138,
+      "step": 1940
+    },
+    {
+      "epoch": 0.20526315789473684,
+      "grad_norm": 0.33088053208710105,
+      "learning_rate": 1e-05,
+      "loss": 0.9892108917236329,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2063157894736842,
+      "grad_norm": 0.2982964440417886,
+      "learning_rate": 1e-05,
+      "loss": 1.0419086456298827,
+      "step": 1960
+    },
+    {
+      "epoch": 0.2073684210526316,
+      "grad_norm": 0.32138828804533626,
+      "learning_rate": 1e-05,
+      "loss": 1.0248214721679687,
+      "step": 1970
+    },
+    {
+      "epoch": 0.20842105263157895,
+      "grad_norm": 0.30789754345974313,
+      "learning_rate": 1e-05,
+      "loss": 1.002777099609375,
+      "step": 1980
+    },
+    {
+      "epoch": 0.2094736842105263,
+      "grad_norm": 0.3270606464425195,
+      "learning_rate": 1e-05,
+      "loss": 1.0345480918884278,
+      "step": 1990
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.29042468729734733,
+      "learning_rate": 1e-05,
+      "loss": 1.0264426231384278,
+      "step": 2000
+    },
+    {
+      "epoch": 0.21157894736842106,
+      "grad_norm": 0.3043191117462647,
+      "learning_rate": 1e-05,
+      "loss": 1.0071794509887695,
+      "step": 2010
+    },
+    {
+      "epoch": 0.21263157894736842,
+      "grad_norm": 0.2943181469204779,
+      "learning_rate": 1e-05,
+      "loss": 1.003769874572754,
+      "step": 2020
+    },
+    {
+      "epoch": 0.21368421052631578,
+      "grad_norm": 0.3516290665830928,
+      "learning_rate": 1e-05,
+      "loss": 1.0497756958007813,
+      "step": 2030
+    },
+    {
+      "epoch": 0.21473684210526317,
+      "grad_norm": 0.3310259631022674,
+      "learning_rate": 1e-05,
+      "loss": 0.9968318939208984,
+      "step": 2040
+    },
+    {
+      "epoch": 0.21578947368421053,
+      "grad_norm": 0.30582035812628394,
+      "learning_rate": 1e-05,
+      "loss": 1.0278396606445312,
+      "step": 2050
+    },
+    {
+      "epoch": 0.2168421052631579,
+      "grad_norm": 0.33426279280979937,
+      "learning_rate": 1e-05,
+      "loss": 1.0261116027832031,
+      "step": 2060
+    },
+    {
+      "epoch": 0.21789473684210525,
+      "grad_norm": 0.36883204086676585,
+      "learning_rate": 1e-05,
+      "loss": 0.9633027076721191,
+      "step": 2070
+    },
+    {
+      "epoch": 0.21894736842105264,
+      "grad_norm": 0.40744719782120126,
+      "learning_rate": 1e-05,
+      "loss": 0.9961132049560547,
+      "step": 2080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.29562210812176204,
+      "learning_rate": 1e-05,
+      "loss": 0.996190071105957,
+      "step": 2090
+    },
+    {
+      "epoch": 0.22105263157894736,
+      "grad_norm": 0.3268581061022905,
+      "learning_rate": 1e-05,
+      "loss": 1.0095386505126953,
+      "step": 2100
+    },
+    {
+      "epoch": 0.22210526315789475,
+      "grad_norm": 0.27675841800164214,
+      "learning_rate": 1e-05,
+      "loss": 0.968992805480957,
+      "step": 2110
+    },
+    {
+      "epoch": 0.2231578947368421,
+      "grad_norm": 0.2945041905650415,
+      "learning_rate": 1e-05,
+      "loss": 0.9765736579895019,
+      "step": 2120
+    },
+    {
+      "epoch": 0.22421052631578947,
+      "grad_norm": 0.3140166554185589,
+      "learning_rate": 1e-05,
+      "loss": 1.011311912536621,
+      "step": 2130
+    },
+    {
+      "epoch": 0.22526315789473683,
+      "grad_norm": 0.2996716689188679,
+      "learning_rate": 1e-05,
+      "loss": 0.9911637306213379,
+      "step": 2140
+    },
+    {
+      "epoch": 0.22631578947368422,
+      "grad_norm": 0.3323037375083453,
+      "learning_rate": 1e-05,
+      "loss": 0.9878484725952148,
+      "step": 2150
+    },
+    {
+      "epoch": 0.22736842105263158,
+      "grad_norm": 0.3118310920021922,
+      "learning_rate": 1e-05,
+      "loss": 0.9956245422363281,
+      "step": 2160
+    },
+    {
+      "epoch": 0.22842105263157894,
+      "grad_norm": 0.29612957623101777,
+      "learning_rate": 1e-05,
+      "loss": 0.9572260856628418,
+      "step": 2170
+    },
+    {
+      "epoch": 0.2294736842105263,
+      "grad_norm": 0.3092177588007431,
+      "learning_rate": 1e-05,
+      "loss": 0.9708539962768554,
+      "step": 2180
+    },
+    {
+      "epoch": 0.2305263157894737,
+      "grad_norm": 0.34097851264513107,
+      "learning_rate": 1e-05,
+      "loss": 0.9612894058227539,
+      "step": 2190
+    },
+    {
+      "epoch": 0.23157894736842105,
+      "grad_norm": 0.3201970765737959,
+      "learning_rate": 1e-05,
+      "loss": 0.9933515548706054,
+      "step": 2200
+    },
+    {
+      "epoch": 0.2326315789473684,
+      "grad_norm": 0.30026412441174655,
+      "learning_rate": 1e-05,
+      "loss": 0.983732795715332,
+      "step": 2210
+    },
+    {
+      "epoch": 0.2336842105263158,
+      "grad_norm": 0.3262421725520795,
+      "learning_rate": 1e-05,
+      "loss": 1.028847599029541,
+      "step": 2220
+    },
+    {
+      "epoch": 0.23473684210526316,
+      "grad_norm": 0.2925521358601703,
+      "learning_rate": 1e-05,
+      "loss": 1.058048915863037,
+      "step": 2230
+    },
+    {
+      "epoch": 0.23578947368421052,
+      "grad_norm": 0.2798259020434305,
+      "learning_rate": 1e-05,
+      "loss": 1.0477144241333007,
+      "step": 2240
+    },
+    {
+      "epoch": 0.23684210526315788,
+      "grad_norm": 0.33744739279624086,
+      "learning_rate": 1e-05,
+      "loss": 1.0694169044494628,
+      "step": 2250
+    },
+    {
+      "epoch": 0.23789473684210527,
+      "grad_norm": 0.30309078001860096,
+      "learning_rate": 1e-05,
+      "loss": 1.1065690994262696,
+      "step": 2260
+    },
+    {
+      "epoch": 0.23894736842105263,
+      "grad_norm": 0.30759355128911176,
+      "learning_rate": 1e-05,
+      "loss": 1.0858617782592774,
+      "step": 2270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2983244672240116,
+      "learning_rate": 1e-05,
+      "loss": 1.0903422355651855,
+      "step": 2280
+    },
+    {
+      "epoch": 0.24105263157894738,
+      "grad_norm": 0.3027858688892284,
+      "learning_rate": 1e-05,
+      "loss": 1.048811912536621,
+      "step": 2290
+    },
+    {
+      "epoch": 0.24210526315789474,
+      "grad_norm": 0.301452811094159,
+      "learning_rate": 1e-05,
+      "loss": 1.0826458930969238,
+      "step": 2300
+    },
+    {
+      "epoch": 0.2431578947368421,
+      "grad_norm": 0.30214576245055397,
+      "learning_rate": 1e-05,
+      "loss": 1.108802032470703,
+      "step": 2310
+    },
+    {
+      "epoch": 0.24421052631578946,
+      "grad_norm": 0.41848961750489755,
+      "learning_rate": 1e-05,
+      "loss": 1.116023635864258,
+      "step": 2320
+    },
+    {
+      "epoch": 0.24526315789473685,
+      "grad_norm": 0.3221801780594187,
+      "learning_rate": 1e-05,
+      "loss": 1.0943788528442382,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2463157894736842,
+      "grad_norm": 0.45861269838021035,
+      "learning_rate": 1e-05,
+      "loss": 1.135487174987793,
+      "step": 2340
+    },
+    {
+      "epoch": 0.24736842105263157,
+      "grad_norm": 0.32165491378716965,
+      "learning_rate": 1e-05,
+      "loss": 1.1145711898803712,
+      "step": 2350
+    },
+    {
+      "epoch": 0.24842105263157896,
+      "grad_norm": 0.333569159910293,
+      "learning_rate": 1e-05,
+      "loss": 1.1319334983825684,
+      "step": 2360
+    },
+    {
+      "epoch": 0.24947368421052632,
+      "grad_norm": 0.33838248449001257,
+      "learning_rate": 1e-05,
+      "loss": 1.1259181022644043,
+      "step": 2370
+    },
+    {
+      "epoch": 0.2505263157894737,
+      "grad_norm": 0.33346574012798813,
+      "learning_rate": 1e-05,
+      "loss": 1.1141367912292481,
+      "step": 2380
+    },
+    {
+      "epoch": 0.25157894736842107,
+      "grad_norm": 0.3276653441464261,
+      "learning_rate": 1e-05,
+      "loss": 1.130275344848633,
+      "step": 2390
+    },
+    {
+      "epoch": 0.25263157894736843,
+      "grad_norm": 0.3192822957537832,
+      "learning_rate": 1e-05,
+      "loss": 1.1037805557250977,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2536842105263158,
+      "grad_norm": 0.3669304174425058,
+      "learning_rate": 1e-05,
+      "loss": 1.1286723136901855,
+      "step": 2410
+    },
+    {
+      "epoch": 0.25473684210526315,
+      "grad_norm": 0.3393904708508052,
+      "learning_rate": 1e-05,
+      "loss": 1.1009031295776368,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2557894736842105,
+      "grad_norm": 0.31853764762246056,
+      "learning_rate": 1e-05,
+      "loss": 1.1165733337402344,
+      "step": 2430
+    },
+    {
+      "epoch": 0.25684210526315787,
+      "grad_norm": 0.3329855329902747,
+      "learning_rate": 1e-05,
+      "loss": 1.1504154205322266,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2578947368421053,
+      "grad_norm": 0.37671348447717584,
+      "learning_rate": 1e-05,
+      "loss": 1.1240503311157226,
+      "step": 2450
+    },
+    {
+      "epoch": 0.25894736842105265,
+      "grad_norm": 0.3421501567356787,
+      "learning_rate": 1e-05,
+      "loss": 1.1019744873046875,
+      "step": 2460
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3002093387417822,
+      "learning_rate": 1e-05,
+      "loss": 1.1313616752624511,
+      "step": 2470
+    },
+    {
+      "epoch": 0.26105263157894737,
+      "grad_norm": 0.32279924016672934,
+      "learning_rate": 1e-05,
+      "loss": 1.1022418975830077,
+      "step": 2480
+    },
+    {
+      "epoch": 0.26210526315789473,
+      "grad_norm": 0.3403423585273364,
+      "learning_rate": 1e-05,
+      "loss": 1.1232851028442383,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.30560185556612735,
+      "learning_rate": 1e-05,
+      "loss": 1.1069017410278321,
+      "step": 2500
+    },
+    {
+      "epoch": 0.26421052631578945,
+      "grad_norm": 0.30763476391149647,
+      "learning_rate": 1e-05,
+      "loss": 1.11282320022583,
+      "step": 2510
+    },
+    {
+      "epoch": 0.26526315789473687,
+      "grad_norm": 0.37240250199581754,
+      "learning_rate": 1e-05,
+      "loss": 1.1273180007934571,
+      "step": 2520
+    },
+    {
+      "epoch": 0.26631578947368423,
+      "grad_norm": 0.3437356791532539,
+      "learning_rate": 1e-05,
+      "loss": 1.1019627571105957,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2673684210526316,
+      "grad_norm": 0.29980048524742753,
+      "learning_rate": 1e-05,
+      "loss": 1.1369990348815917,
+      "step": 2540
+    },
+    {
+      "epoch": 0.26842105263157895,
+      "grad_norm": 0.32154441119697535,
+      "learning_rate": 1e-05,
+      "loss": 1.1139318466186523,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2694736842105263,
+      "grad_norm": 0.33610147541420793,
+      "learning_rate": 1e-05,
+      "loss": 1.1036800384521483,
+      "step": 2560
+    },
+    {
+      "epoch": 0.27052631578947367,
+      "grad_norm": 0.30362643557793717,
+      "learning_rate": 1e-05,
+      "loss": 1.0797775268554688,
+      "step": 2570
+    },
+    {
+      "epoch": 0.27157894736842103,
+      "grad_norm": 0.2877924151250326,
+      "learning_rate": 1e-05,
+      "loss": 1.080092430114746,
+      "step": 2580
+    },
+    {
+      "epoch": 0.27263157894736845,
+      "grad_norm": 0.29869804143826023,
+      "learning_rate": 1e-05,
+      "loss": 1.077783203125,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2736842105263158,
+      "grad_norm": 0.34473542208687236,
+      "learning_rate": 1e-05,
+      "loss": 1.0669015884399413,
+      "step": 2600
+    },
+    {
+      "epoch": 0.27473684210526317,
+      "grad_norm": 0.3581822660700594,
+      "learning_rate": 1e-05,
+      "loss": 1.063136100769043,
+      "step": 2610
+    },
+    {
+      "epoch": 0.27578947368421053,
+      "grad_norm": 0.3356212756290226,
+      "learning_rate": 1e-05,
+      "loss": 1.0607318878173828,
+      "step": 2620
+    },
+    {
+      "epoch": 0.2768421052631579,
+      "grad_norm": 0.3493096221223222,
+      "learning_rate": 1e-05,
+      "loss": 1.0549909591674804,
+      "step": 2630
+    },
+    {
+      "epoch": 0.27789473684210525,
+      "grad_norm": 0.2867518223853459,
+      "learning_rate": 1e-05,
+      "loss": 1.0098052024841309,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2789473684210526,
+      "grad_norm": 0.29105579220967664,
+      "learning_rate": 1e-05,
+      "loss": 1.0149802207946776,
+      "step": 2650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3154016131023938,
+      "learning_rate": 1e-05,
+      "loss": 1.0352468490600586,
+      "step": 2660
+    },
+    {
+      "epoch": 0.2810526315789474,
+      "grad_norm": 0.2837117539607667,
+      "learning_rate": 1e-05,
+      "loss": 0.9905804634094239,
+      "step": 2670
+    },
+    {
+      "epoch": 0.28210526315789475,
+      "grad_norm": 0.31860373920142865,
+      "learning_rate": 1e-05,
+      "loss": 1.0011009216308593,
+      "step": 2680
+    },
+    {
+      "epoch": 0.2831578947368421,
+      "grad_norm": 0.2893619742741659,
+      "learning_rate": 1e-05,
+      "loss": 1.0173326492309571,
+      "step": 2690
+    },
+    {
+      "epoch": 0.28421052631578947,
+      "grad_norm": 0.270702675184311,
+      "learning_rate": 1e-05,
+      "loss": 0.9944775581359864,
+      "step": 2700
+    },
+    {
+      "epoch": 0.28526315789473683,
+      "grad_norm": 0.3045819565820902,
+      "learning_rate": 1e-05,
+      "loss": 0.9745292663574219,
+      "step": 2710
+    },
+    {
+      "epoch": 0.2863157894736842,
+      "grad_norm": 0.2765459748132974,
+      "learning_rate": 1e-05,
+      "loss": 0.977507209777832,
+      "step": 2720
+    },
+    {
+      "epoch": 0.2873684210526316,
+      "grad_norm": 0.3226076872319736,
+      "learning_rate": 1e-05,
+      "loss": 0.9664097785949707,
+      "step": 2730
+    },
+    {
+      "epoch": 0.28842105263157897,
+      "grad_norm": 0.28848427026842294,
+      "learning_rate": 1e-05,
+      "loss": 0.9569656372070312,
+      "step": 2740
+    },
+    {
+      "epoch": 0.2894736842105263,
+      "grad_norm": 0.3088824463396163,
+      "learning_rate": 1e-05,
+      "loss": 0.9399412155151368,
+      "step": 2750
+    },
+    {
+      "epoch": 0.2905263157894737,
+      "grad_norm": 0.2929952035675146,
+      "learning_rate": 1e-05,
+      "loss": 0.938800048828125,
+      "step": 2760
+    },
+    {
+      "epoch": 0.29157894736842105,
+      "grad_norm": 0.2764593332475593,
+      "learning_rate": 1e-05,
+      "loss": 0.9413426399230957,
+      "step": 2770
+    },
+    {
+      "epoch": 0.2926315789473684,
+      "grad_norm": 0.26496019314398783,
+      "learning_rate": 1e-05,
+      "loss": 0.9343229293823242,
+      "step": 2780
+    },
+    {
+      "epoch": 0.29368421052631577,
+      "grad_norm": 0.3315344046024685,
+      "learning_rate": 1e-05,
+      "loss": 0.9172769546508789,
+      "step": 2790
+    },
+    {
+      "epoch": 0.29473684210526313,
+      "grad_norm": 0.25046703382655816,
+      "learning_rate": 1e-05,
+      "loss": 0.9195137023925781,
+      "step": 2800
+    },
+    {
+      "epoch": 0.29578947368421055,
+      "grad_norm": 0.2814003479977443,
+      "learning_rate": 1e-05,
+      "loss": 0.9100202560424805,
+      "step": 2810
+    },
+    {
+      "epoch": 0.2968421052631579,
+      "grad_norm": 0.2772777526441368,
+      "learning_rate": 1e-05,
+      "loss": 0.9039760589599609,
+      "step": 2820
+    },
+    {
+      "epoch": 0.29789473684210527,
+      "grad_norm": 0.2694289881211397,
+      "learning_rate": 1e-05,
+      "loss": 0.8933679580688476,
+      "step": 2830
+    },
+    {
+      "epoch": 0.29894736842105263,
+      "grad_norm": 0.3109222728434967,
+      "learning_rate": 1e-05,
+      "loss": 0.8730718612670898,
+      "step": 2840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28321721094254454,
+      "learning_rate": 1e-05,
+      "loss": 0.901797103881836,
+      "step": 2850
+    },
+    {
+      "epoch": 0.30105263157894735,
+      "grad_norm": 0.25125105120265556,
+      "learning_rate": 1e-05,
+      "loss": 0.8691996574401856,
+      "step": 2860
+    },
+    {
+      "epoch": 0.3021052631578947,
+      "grad_norm": 0.2764056197593486,
+      "learning_rate": 1e-05,
+      "loss": 0.8760793685913086,
+      "step": 2870
+    },
+    {
+      "epoch": 0.3031578947368421,
+      "grad_norm": 0.2511519144620448,
+      "learning_rate": 1e-05,
+      "loss": 0.8862150192260743,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3042105263157895,
+      "grad_norm": 0.2620345695952759,
+      "learning_rate": 1e-05,
+      "loss": 0.8431035995483398,
+      "step": 2890
+    },
+    {
+      "epoch": 0.30526315789473685,
+      "grad_norm": 0.25677875613640994,
+      "learning_rate": 1e-05,
+      "loss": 0.851867389678955,
+      "step": 2900
+    },
+    {
+      "epoch": 0.3063157894736842,
+      "grad_norm": 0.25970331631247395,
+      "learning_rate": 1e-05,
+      "loss": 0.8187185287475586,
+      "step": 2910
+    },
+    {
+      "epoch": 0.30736842105263157,
+      "grad_norm": 0.25785799539820536,
+      "learning_rate": 1e-05,
+      "loss": 0.8182131767272949,
+      "step": 2920
+    },
+    {
+      "epoch": 0.30842105263157893,
+      "grad_norm": 0.2448949752429678,
+      "learning_rate": 1e-05,
+      "loss": 0.8069913864135743,
+      "step": 2930
+    },
+    {
+      "epoch": 0.3094736842105263,
+      "grad_norm": 0.25994184955797645,
+      "learning_rate": 1e-05,
+      "loss": 0.8174491882324219,
+      "step": 2940
+    },
+    {
+      "epoch": 0.3105263157894737,
+      "grad_norm": 0.25254881331995876,
+      "learning_rate": 1e-05,
+      "loss": 0.7967897415161133,
+      "step": 2950
+    },
+    {
+      "epoch": 0.31157894736842107,
+      "grad_norm": 0.28391323026283616,
+      "learning_rate": 1e-05,
+      "loss": 0.7990935802459717,
+      "step": 2960
+    },
+    {
+      "epoch": 0.3126315789473684,
+      "grad_norm": 0.43343020869129395,
+      "learning_rate": 1e-05,
+      "loss": 0.8045951843261718,
+      "step": 2970
+    },
+    {
+      "epoch": 0.3136842105263158,
+      "grad_norm": 0.2582072682523002,
+      "learning_rate": 1e-05,
+      "loss": 0.7698857307434082,
+      "step": 2980
+    },
+    {
+      "epoch": 0.31473684210526315,
+      "grad_norm": 0.243598864836983,
+      "learning_rate": 1e-05,
+      "loss": 0.7765778064727783,
+      "step": 2990
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.2582370009128698,
+      "learning_rate": 1e-05,
+      "loss": 0.8456914901733399,
+      "step": 3000
+    },
+    {
+      "epoch": 0.31684210526315787,
+      "grad_norm": 0.26626685243032383,
+      "learning_rate": 1e-05,
+      "loss": 0.855282211303711,
+      "step": 3010
+    },
+    {
+      "epoch": 0.3178947368421053,
+      "grad_norm": 0.2310589670131793,
+      "learning_rate": 1e-05,
+      "loss": 0.859808349609375,
+      "step": 3020
+    },
+    {
+      "epoch": 0.31894736842105265,
+      "grad_norm": 0.27335442527107423,
+      "learning_rate": 1e-05,
+      "loss": 0.9612586975097657,
+      "step": 3030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3065007246994013,
+      "learning_rate": 1e-05,
+      "loss": 0.901555061340332,
+      "step": 3040
+    },
+    {
+      "epoch": 0.32105263157894737,
+      "grad_norm": 0.28770966734601294,
+      "learning_rate": 1e-05,
+      "loss": 0.9473600387573242,
+      "step": 3050
+    },
+    {
+      "epoch": 0.32210526315789473,
+      "grad_norm": 0.2752232282340422,
+      "learning_rate": 1e-05,
+      "loss": 0.9013284683227539,
+      "step": 3060
+    },
+    {
+      "epoch": 0.3231578947368421,
+      "grad_norm": 0.2613558249820635,
+      "learning_rate": 1e-05,
+      "loss": 0.941219711303711,
+      "step": 3070
+    },
+    {
+      "epoch": 0.32421052631578945,
+      "grad_norm": 0.2572150590264271,
+      "learning_rate": 1e-05,
+      "loss": 0.8851035118103028,
+      "step": 3080
+    },
+    {
+      "epoch": 0.32526315789473687,
+      "grad_norm": 0.2659592304438082,
+      "learning_rate": 1e-05,
+      "loss": 0.8896900177001953,
+      "step": 3090
+    },
+    {
+      "epoch": 0.3263157894736842,
+      "grad_norm": 0.27895819447274395,
+      "learning_rate": 1e-05,
+      "loss": 0.9277690887451172,
+      "step": 3100
+    },
+    {
+      "epoch": 0.3273684210526316,
+      "grad_norm": 0.2641713980374385,
+      "learning_rate": 1e-05,
+      "loss": 0.8948198318481445,
+      "step": 3110
+    },
+    {
+      "epoch": 0.32842105263157895,
+      "grad_norm": 0.2479118013526724,
+      "learning_rate": 1e-05,
+      "loss": 0.9414800643920899,
+      "step": 3120
+    },
+    {
+      "epoch": 0.3294736842105263,
+      "grad_norm": 0.27870010746583124,
+      "learning_rate": 1e-05,
+      "loss": 0.907012939453125,
+      "step": 3130
+    },
+    {
+      "epoch": 0.33052631578947367,
+      "grad_norm": 0.26105014568654633,
+      "learning_rate": 1e-05,
+      "loss": 0.9367162704467773,
+      "step": 3140
+    },
+    {
+      "epoch": 0.33157894736842103,
+      "grad_norm": 0.2924788907841983,
+      "learning_rate": 1e-05,
+      "loss": 0.9563218116760254,
+      "step": 3150
+    },
+    {
+      "epoch": 0.33263157894736844,
+      "grad_norm": 0.28255839478009326,
+      "learning_rate": 1e-05,
+      "loss": 0.8844303131103516,
+      "step": 3160
+    },
+    {
+      "epoch": 0.3336842105263158,
+      "grad_norm": 0.2828532905450654,
+      "learning_rate": 1e-05,
+      "loss": 0.9386220932006836,
+      "step": 3170
+    },
+    {
+      "epoch": 0.33473684210526317,
+      "grad_norm": 0.25401117594706496,
+      "learning_rate": 1e-05,
+      "loss": 0.9213560104370118,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3357894736842105,
+      "grad_norm": 0.2501930071990557,
+      "learning_rate": 1e-05,
+      "loss": 0.9632314682006836,
+      "step": 3190
+    },
+    {
+      "epoch": 0.3368421052631579,
+      "grad_norm": 0.30472767085559477,
+      "learning_rate": 1e-05,
+      "loss": 0.9519287109375,
+      "step": 3200
+    },
+    {
+      "epoch": 0.33789473684210525,
+      "grad_norm": 0.26341827629621384,
+      "learning_rate": 1e-05,
+      "loss": 0.9274541854858398,
+      "step": 3210
+    },
+    {
+      "epoch": 0.3389473684210526,
+      "grad_norm": 0.2764205690430962,
+      "learning_rate": 1e-05,
+      "loss": 0.9327966690063476,
+      "step": 3220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25147417115020193,
+      "learning_rate": 1e-05,
+      "loss": 0.9424642562866211,
+      "step": 3230
+    },
+    {
+      "epoch": 0.3410526315789474,
+      "grad_norm": 0.26043119517603247,
+      "learning_rate": 1e-05,
+      "loss": 0.9493393898010254,
+      "step": 3240
+    },
+    {
+      "epoch": 0.34210526315789475,
+      "grad_norm": 0.6110150766465257,
+      "learning_rate": 1e-05,
+      "loss": 0.9394001007080078,
+      "step": 3250
+    },
+    {
+      "epoch": 0.3431578947368421,
+      "grad_norm": 0.27303782768572793,
+      "learning_rate": 1e-05,
+      "loss": 0.9700277328491211,
+      "step": 3260
+    },
+    {
+      "epoch": 0.34421052631578947,
+      "grad_norm": 0.24669196245541353,
+      "learning_rate": 1e-05,
+      "loss": 0.9622254371643066,
+      "step": 3270
+    },
+    {
+      "epoch": 0.3452631578947368,
+      "grad_norm": 0.26404708621141576,
+      "learning_rate": 1e-05,
+      "loss": 0.9377164840698242,
+      "step": 3280
+    },
+    {
+      "epoch": 0.3463157894736842,
+      "grad_norm": 0.2665641779619641,
+      "learning_rate": 1e-05,
+      "loss": 0.9563883781433106,
+      "step": 3290
+    },
+    {
+      "epoch": 0.3473684210526316,
+      "grad_norm": 0.25499003972533213,
+      "learning_rate": 1e-05,
+      "loss": 0.9610318183898926,
+      "step": 3300
+    },
+    {
+      "epoch": 0.34842105263157896,
+      "grad_norm": 0.3843332129598115,
+      "learning_rate": 1e-05,
+      "loss": 0.9404146194458007,
+      "step": 3310
+    },
+    {
+      "epoch": 0.3494736842105263,
+      "grad_norm": 0.27145690406705003,
+      "learning_rate": 1e-05,
+      "loss": 0.94205322265625,
+      "step": 3320
+    },
+    {
+      "epoch": 0.3505263157894737,
+      "grad_norm": 0.28177107781037725,
+      "learning_rate": 1e-05,
+      "loss": 0.9393021583557128,
+      "step": 3330
+    },
+    {
+      "epoch": 0.35157894736842105,
+      "grad_norm": 0.24068981637300718,
+      "learning_rate": 1e-05,
+      "loss": 0.9515616416931152,
+      "step": 3340
+    },
+    {
+      "epoch": 0.3526315789473684,
+      "grad_norm": 0.2946379216676235,
+      "learning_rate": 1e-05,
+      "loss": 0.9463732719421387,
+      "step": 3350
+    },
+    {
+      "epoch": 0.35368421052631577,
+      "grad_norm": 0.264987633317566,
+      "learning_rate": 1e-05,
+      "loss": 0.9331771850585937,
+      "step": 3360
+    },
+    {
+      "epoch": 0.3547368421052632,
+      "grad_norm": 0.2581747007060754,
+      "learning_rate": 1e-05,
+      "loss": 0.9378446578979492,
+      "step": 3370
+    },
+    {
+      "epoch": 0.35578947368421054,
+      "grad_norm": 0.2797952282216625,
+      "learning_rate": 1e-05,
+      "loss": 0.9576354026794434,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3568421052631579,
+      "grad_norm": 0.261681085461167,
+      "learning_rate": 1e-05,
+      "loss": 0.965583324432373,
+      "step": 3390
+    },
+    {
+      "epoch": 0.35789473684210527,
+      "grad_norm": 0.337303903208587,
+      "learning_rate": 1e-05,
+      "loss": 0.9428577423095703,
+      "step": 3400
+    },
+    {
+      "epoch": 0.3589473684210526,
+      "grad_norm": 0.30655372847000906,
+      "learning_rate": 1e-05,
+      "loss": 0.9728082656860352,
+      "step": 3410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.253761196348214,
+      "learning_rate": 1e-05,
+      "loss": 0.9373969078063965,
+      "step": 3420
+    },
+    {
+      "epoch": 0.36105263157894735,
+      "grad_norm": 0.48315234808872015,
+      "learning_rate": 1e-05,
+      "loss": 0.9547054290771484,
+      "step": 3430
+    },
+    {
+      "epoch": 0.36210526315789476,
+      "grad_norm": 0.24875568640404988,
+      "learning_rate": 1e-05,
+      "loss": 0.9445707321166992,
+      "step": 3440
+    },
+    {
+      "epoch": 0.3631578947368421,
+      "grad_norm": 0.25871599876118345,
+      "learning_rate": 1e-05,
+      "loss": 0.9430584907531738,
+      "step": 3450
+    },
+    {
+      "epoch": 0.3642105263157895,
+      "grad_norm": 0.26556147267664815,
+      "learning_rate": 1e-05,
+      "loss": 0.9575901031494141,
+      "step": 3460
+    },
+    {
+      "epoch": 0.36526315789473685,
+      "grad_norm": 0.2643779108127813,
+      "learning_rate": 1e-05,
+      "loss": 0.923575496673584,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3663157894736842,
+      "grad_norm": 0.2686310919769273,
+      "learning_rate": 1e-05,
+      "loss": 0.9519380569458008,
+      "step": 3480
+    },
+    {
+      "epoch": 0.36736842105263157,
+      "grad_norm": 0.24713311191799517,
+      "learning_rate": 1e-05,
+      "loss": 0.9373991012573242,
+      "step": 3490
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 0.2648795731227103,
+      "learning_rate": 1e-05,
+      "loss": 0.953524398803711,
+      "step": 3500
+    },
+    {
+      "epoch": 0.36947368421052634,
+      "grad_norm": 0.250598394022935,
+      "learning_rate": 1e-05,
+      "loss": 0.9408896446228028,
+      "step": 3510
+    },
+    {
+      "epoch": 0.3705263157894737,
+      "grad_norm": 0.2607097300807116,
+      "learning_rate": 1e-05,
+      "loss": 0.946784782409668,
+      "step": 3520
+    },
+    {
+      "epoch": 0.37157894736842106,
+      "grad_norm": 0.2495077706110475,
+      "learning_rate": 1e-05,
+      "loss": 0.9516042709350586,
+      "step": 3530
+    },
+    {
+      "epoch": 0.3726315789473684,
+      "grad_norm": 0.2615572932379115,
+      "learning_rate": 1e-05,
+      "loss": 0.9414764404296875,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3736842105263158,
+      "grad_norm": 0.274761768880261,
+      "learning_rate": 1e-05,
+      "loss": 0.9447465896606445,
+      "step": 3550
+    },
+    {
+      "epoch": 0.37473684210526315,
+      "grad_norm": 0.2646123382169523,
+      "learning_rate": 1e-05,
+      "loss": 0.9470996856689453,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3757894736842105,
+      "grad_norm": 0.25864111591843764,
+      "learning_rate": 1e-05,
+      "loss": 0.9418252944946289,
+      "step": 3570
+    },
+    {
+      "epoch": 0.37684210526315787,
+      "grad_norm": 0.35627655681022075,
+      "learning_rate": 1e-05,
+      "loss": 0.9234872817993164,
+      "step": 3580
+    },
+    {
+      "epoch": 0.3778947368421053,
+      "grad_norm": 0.24107375114754578,
+      "learning_rate": 1e-05,
+      "loss": 0.9505449295043945,
+      "step": 3590
+    },
+    {
+      "epoch": 0.37894736842105264,
+      "grad_norm": 0.7938775576829187,
+      "learning_rate": 1e-05,
+      "loss": 0.9380820274353028,
+      "step": 3600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24790991372094248,
+      "learning_rate": 1e-05,
+      "loss": 0.9071937561035156,
+      "step": 3610
+    },
+    {
+      "epoch": 0.38105263157894737,
+      "grad_norm": 0.2672917016981947,
+      "learning_rate": 1e-05,
+      "loss": 0.9715073585510254,
+      "step": 3620
+    },
+    {
+      "epoch": 0.3821052631578947,
+      "grad_norm": 0.2601576405680808,
+      "learning_rate": 1e-05,
+      "loss": 0.9504000663757324,
+      "step": 3630
+    },
+    {
+      "epoch": 0.3831578947368421,
+      "grad_norm": 0.2726887361839276,
+      "learning_rate": 1e-05,
+      "loss": 0.9281610488891602,
+      "step": 3640
+    },
+    {
+      "epoch": 0.38421052631578945,
+      "grad_norm": 0.26282423899172913,
+      "learning_rate": 1e-05,
+      "loss": 0.9401857376098632,
+      "step": 3650
+    },
+    {
+      "epoch": 0.38526315789473686,
+      "grad_norm": 0.29526796956244605,
+      "learning_rate": 1e-05,
+      "loss": 0.9588068008422852,
+      "step": 3660
+    },
+    {
+      "epoch": 0.3863157894736842,
+      "grad_norm": 0.2686700397783435,
+      "learning_rate": 1e-05,
+      "loss": 0.9346155166625977,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3873684210526316,
+      "grad_norm": 0.25809483768949126,
+      "learning_rate": 1e-05,
+      "loss": 0.933635139465332,
+      "step": 3680
+    },
+    {
+      "epoch": 0.38842105263157894,
+      "grad_norm": 0.32035948860633023,
+      "learning_rate": 1e-05,
+      "loss": 0.9468857765197753,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3894736842105263,
+      "grad_norm": 0.2581304235576142,
+      "learning_rate": 1e-05,
+      "loss": 0.9245304107666016,
+      "step": 3700
+    },
+    {
+      "epoch": 0.39052631578947367,
+      "grad_norm": 0.25635353674508027,
+      "learning_rate": 1e-05,
+      "loss": 0.9250963211059571,
+      "step": 3710
+    },
+    {
+      "epoch": 0.391578947368421,
+      "grad_norm": 0.30079112444552863,
+      "learning_rate": 1e-05,
+      "loss": 0.9293127059936523,
+      "step": 3720
+    },
+    {
+      "epoch": 0.39263157894736844,
+      "grad_norm": 0.26872748518718403,
+      "learning_rate": 1e-05,
+      "loss": 0.9212482452392579,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3936842105263158,
+      "grad_norm": 0.2484685941151488,
+      "learning_rate": 1e-05,
+      "loss": 0.960307502746582,
+      "step": 3740
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "grad_norm": 0.27344877657502253,
+      "learning_rate": 1e-05,
+      "loss": 0.9275402069091797,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3957894736842105,
+      "grad_norm": 0.2456452369171624,
+      "learning_rate": 1e-05,
+      "loss": 0.9440584182739258,
+      "step": 3760
+    },
+    {
+      "epoch": 0.3968421052631579,
+      "grad_norm": 0.27459477277718214,
+      "learning_rate": 1e-05,
+      "loss": 0.9314395904541015,
+      "step": 3770
+    },
+    {
+      "epoch": 0.39789473684210525,
+      "grad_norm": 0.5048503189193233,
+      "learning_rate": 1e-05,
+      "loss": 0.9263954162597656,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3989473684210526,
+      "grad_norm": 0.24749430027671118,
+      "learning_rate": 1e-05,
+      "loss": 0.9334461212158203,
+      "step": 3790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2682971056740341,
+      "learning_rate": 1e-05,
+      "loss": 0.9248899459838867,
+      "step": 3800
+    },
+    {
+      "epoch": 0.4010526315789474,
+      "grad_norm": 0.24936953355959884,
+      "learning_rate": 1e-05,
+      "loss": 0.9108712196350097,
+      "step": 3810
+    },
+    {
+      "epoch": 0.40210526315789474,
+      "grad_norm": 0.28362448310038174,
+      "learning_rate": 1e-05,
+      "loss": 0.9368877410888672,
+      "step": 3820
+    },
+    {
+      "epoch": 0.4031578947368421,
+      "grad_norm": 0.22780868673903557,
+      "learning_rate": 1e-05,
+      "loss": 0.9117192268371582,
+      "step": 3830
+    },
+    {
+      "epoch": 0.40421052631578946,
+      "grad_norm": 0.24670625810295413,
+      "learning_rate": 1e-05,
+      "loss": 0.9132587432861328,
+      "step": 3840
+    },
+    {
+      "epoch": 0.4052631578947368,
+      "grad_norm": 0.2637667542747437,
+      "learning_rate": 1e-05,
+      "loss": 0.9173944473266602,
+      "step": 3850
+    },
+    {
+      "epoch": 0.4063157894736842,
+      "grad_norm": 0.24882603363207853,
+      "learning_rate": 1e-05,
+      "loss": 0.9187602996826172,
+      "step": 3860
+    },
+    {
+      "epoch": 0.4073684210526316,
+      "grad_norm": 0.30568929291706987,
+      "learning_rate": 1e-05,
+      "loss": 0.9228479385375976,
+      "step": 3870
+    },
+    {
+      "epoch": 0.40842105263157896,
+      "grad_norm": 0.25010478244787504,
+      "learning_rate": 1e-05,
+      "loss": 0.9254318237304687,
+      "step": 3880
+    },
+    {
+      "epoch": 0.4094736842105263,
+      "grad_norm": 0.24286121303600974,
+      "learning_rate": 1e-05,
+      "loss": 0.9048201560974121,
+      "step": 3890
+    },
+    {
+      "epoch": 0.4105263157894737,
+      "grad_norm": 0.2505952841786791,
+      "learning_rate": 1e-05,
+      "loss": 0.9247965812683105,
+      "step": 3900
+    },
+    {
+      "epoch": 0.41157894736842104,
+      "grad_norm": 0.24770181475412464,
+      "learning_rate": 1e-05,
+      "loss": 0.9100747108459473,
+      "step": 3910
+    },
+    {
+      "epoch": 0.4126315789473684,
+      "grad_norm": 0.2746177026063864,
+      "learning_rate": 1e-05,
+      "loss": 0.9113056182861328,
+      "step": 3920
+    },
+    {
+      "epoch": 0.41368421052631577,
+      "grad_norm": 0.24405092258104175,
+      "learning_rate": 1e-05,
+      "loss": 0.9152275085449219,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4147368421052632,
+      "grad_norm": 0.29305350106170663,
+      "learning_rate": 1e-05,
+      "loss": 0.9050154685974121,
+      "step": 3940
+    },
+    {
+      "epoch": 0.41578947368421054,
+      "grad_norm": 0.26434523054258896,
+      "learning_rate": 1e-05,
+      "loss": 0.9207850456237793,
+      "step": 3950
+    },
+    {
+      "epoch": 0.4168421052631579,
+      "grad_norm": 0.29837325445139984,
+      "learning_rate": 1e-05,
+      "loss": 0.9388660430908203,
+      "step": 3960
+    },
+    {
+      "epoch": 0.41789473684210526,
+      "grad_norm": 0.2717239076259495,
+      "learning_rate": 1e-05,
+      "loss": 0.9400924682617188,
+      "step": 3970
+    },
+    {
+      "epoch": 0.4189473684210526,
+      "grad_norm": 0.26337599915514437,
+      "learning_rate": 1e-05,
+      "loss": 0.9243921279907227,
+      "step": 3980
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.24642873132492987,
+      "learning_rate": 1e-05,
+      "loss": 0.9089398384094238,
+      "step": 3990
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.27090930844967254,
+      "learning_rate": 1e-05,
+      "loss": 0.9091521263122558,
+      "step": 4000
+    },
+    {
+      "epoch": 0.42210526315789476,
+      "grad_norm": 0.2393293823462255,
+      "learning_rate": 1e-05,
+      "loss": 0.9174625396728515,
+      "step": 4010
+    },
+    {
+      "epoch": 0.4231578947368421,
+      "grad_norm": 0.4078832761028804,
+      "learning_rate": 1e-05,
+      "loss": 0.927030372619629,
+      "step": 4020
+    },
+    {
+      "epoch": 0.4242105263157895,
+      "grad_norm": 0.2698088881161566,
+      "learning_rate": 1e-05,
+      "loss": 0.9080303192138672,
+      "step": 4030
+    },
+    {
+      "epoch": 0.42526315789473684,
+      "grad_norm": 0.255325737866685,
+      "learning_rate": 1e-05,
+      "loss": 0.9064571380615234,
+      "step": 4040
+    },
+    {
+      "epoch": 0.4263157894736842,
+      "grad_norm": 0.2568312242034188,
+      "learning_rate": 1e-05,
+      "loss": 0.9298416137695312,
+      "step": 4050
+    },
+    {
+      "epoch": 0.42736842105263156,
+      "grad_norm": 0.26187662506631243,
+      "learning_rate": 1e-05,
+      "loss": 0.9239648818969727,
+      "step": 4060
+    },
+    {
+      "epoch": 0.4284210526315789,
+      "grad_norm": 0.2494094098589495,
+      "learning_rate": 1e-05,
+      "loss": 0.9692612648010254,
+      "step": 4070
+    },
+    {
+      "epoch": 0.42947368421052634,
+      "grad_norm": 0.28909665785845545,
+      "learning_rate": 1e-05,
+      "loss": 0.9519695281982422,
+      "step": 4080
+    },
+    {
+      "epoch": 0.4305263157894737,
+      "grad_norm": 0.28717211323885666,
+      "learning_rate": 1e-05,
+      "loss": 1.0089866638183593,
+      "step": 4090
+    },
+    {
+      "epoch": 0.43157894736842106,
+      "grad_norm": 0.2862389042051491,
+      "learning_rate": 1e-05,
+      "loss": 1.0147268295288085,
+      "step": 4100
+    },
+    {
+      "epoch": 0.4326315789473684,
+      "grad_norm": 0.27674719064676273,
+      "learning_rate": 1e-05,
+      "loss": 1.0389043807983398,
+      "step": 4110
+    },
+    {
+      "epoch": 0.4336842105263158,
+      "grad_norm": 0.28137239306683065,
+      "learning_rate": 1e-05,
+      "loss": 1.0330793380737304,
+      "step": 4120
+    },
+    {
+      "epoch": 0.43473684210526314,
+      "grad_norm": 0.2653791515371348,
+      "learning_rate": 1e-05,
+      "loss": 1.058261013031006,
+      "step": 4130
+    },
+    {
+      "epoch": 0.4357894736842105,
+      "grad_norm": 0.4352221130790731,
+      "learning_rate": 1e-05,
+      "loss": 1.0831575393676758,
+      "step": 4140
+    },
+    {
+      "epoch": 0.4368421052631579,
+      "grad_norm": 0.34717273132813925,
+      "learning_rate": 1e-05,
+      "loss": 1.1029497146606446,
+      "step": 4150
+    },
+    {
+      "epoch": 0.4378947368421053,
+      "grad_norm": 0.4350384981308931,
+      "learning_rate": 1e-05,
+      "loss": 1.1080530166625977,
+      "step": 4160
+    },
+    {
+      "epoch": 0.43894736842105264,
+      "grad_norm": 0.2702920147621635,
+      "learning_rate": 1e-05,
+      "loss": 1.0952021598815918,
+      "step": 4170
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32518677496182863,
+      "learning_rate": 1e-05,
+      "loss": 1.1344825744628906,
+      "step": 4180
+    },
+    {
+      "epoch": 0.44105263157894736,
+      "grad_norm": 0.2705884864642059,
+      "learning_rate": 1e-05,
+      "loss": 1.1103446006774902,
+      "step": 4190
+    },
+    {
+      "epoch": 0.4421052631578947,
+      "grad_norm": 0.36091907065286305,
+      "learning_rate": 1e-05,
+      "loss": 1.1062320709228515,
+      "step": 4200
+    },
+    {
+      "epoch": 0.4431578947368421,
+      "grad_norm": 0.35738900500632936,
+      "learning_rate": 1e-05,
+      "loss": 1.1368200302124023,
+      "step": 4210
+    },
+    {
+      "epoch": 0.4442105263157895,
+      "grad_norm": 0.40017754002961986,
+      "learning_rate": 1e-05,
+      "loss": 1.1405187606811524,
+      "step": 4220
+    },
+    {
+      "epoch": 0.44526315789473686,
+      "grad_norm": 0.29937744451349657,
+      "learning_rate": 1e-05,
+      "loss": 1.1330222129821776,
+      "step": 4230
+    },
+    {
+      "epoch": 0.4463157894736842,
+      "grad_norm": 0.31033840841974264,
+      "learning_rate": 1e-05,
+      "loss": 1.1080530166625977,
+      "step": 4240
+    },
+    {
+      "epoch": 0.4473684210526316,
+      "grad_norm": 0.3427680541820844,
+      "learning_rate": 1e-05,
+      "loss": 1.156318473815918,
+      "step": 4250
+    },
+    {
+      "epoch": 0.44842105263157894,
+      "grad_norm": 0.3270771641370226,
+      "learning_rate": 1e-05,
+      "loss": 1.1515482902526855,
+      "step": 4260
+    },
+    {
+      "epoch": 0.4494736842105263,
+      "grad_norm": 0.3491740451141998,
+      "learning_rate": 1e-05,
+      "loss": 1.1680402755737305,
+      "step": 4270
+    },
+    {
+      "epoch": 0.45052631578947366,
+      "grad_norm": 0.2922884151519592,
+      "learning_rate": 1e-05,
+      "loss": 1.1584228515625,
+      "step": 4280
+    },
+    {
+      "epoch": 0.4515789473684211,
+      "grad_norm": 0.2812627012410352,
+      "learning_rate": 1e-05,
+      "loss": 1.1474220275878906,
+      "step": 4290
+    },
+    {
+      "epoch": 0.45263157894736844,
+      "grad_norm": 0.3140712414228219,
+      "learning_rate": 1e-05,
+      "loss": 1.1687448501586915,
+      "step": 4300
+    },
+    {
+      "epoch": 0.4536842105263158,
+      "grad_norm": 0.2874768101796482,
+      "learning_rate": 1e-05,
+      "loss": 1.1401483535766601,
+      "step": 4310
+    },
+    {
+      "epoch": 0.45473684210526316,
+      "grad_norm": 0.35810073701807726,
+      "learning_rate": 1e-05,
+      "loss": 1.118852424621582,
+      "step": 4320
+    },
+    {
+      "epoch": 0.4557894736842105,
+      "grad_norm": 0.3156073695905802,
+      "learning_rate": 1e-05,
+      "loss": 1.149275016784668,
+      "step": 4330
+    },
+    {
+      "epoch": 0.4568421052631579,
+      "grad_norm": 0.3220341892442449,
+      "learning_rate": 1e-05,
+      "loss": 1.1523815155029298,
+      "step": 4340
+    },
+    {
+      "epoch": 0.45789473684210524,
+      "grad_norm": 0.31739018941379293,
+      "learning_rate": 1e-05,
+      "loss": 1.173818588256836,
+      "step": 4350
+    },
+    {
+      "epoch": 0.4589473684210526,
+      "grad_norm": 0.2728064507149881,
+      "learning_rate": 1e-05,
+      "loss": 1.1483546257019044,
+      "step": 4360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29744707047011326,
+      "learning_rate": 1e-05,
+      "loss": 1.1139816284179687,
+      "step": 4370
+    },
+    {
+      "epoch": 0.4610526315789474,
+      "grad_norm": 0.29675632353889564,
+      "learning_rate": 1e-05,
+      "loss": 1.1581433296203614,
+      "step": 4380
+    },
+    {
+      "epoch": 0.46210526315789474,
+      "grad_norm": 0.3215771461742891,
+      "learning_rate": 1e-05,
+      "loss": 1.1533740997314452,
+      "step": 4390
+    },
+    {
+      "epoch": 0.4631578947368421,
+      "grad_norm": 0.3864992533344979,
+      "learning_rate": 1e-05,
+      "loss": 1.1895965576171874,
+      "step": 4400
+    },
+    {
+      "epoch": 0.46421052631578946,
+      "grad_norm": 0.38025844152693583,
+      "learning_rate": 1e-05,
+      "loss": 1.1792179107666017,
+      "step": 4410
+    },
+    {
+      "epoch": 0.4652631578947368,
+      "grad_norm": 0.27180937662729865,
+      "learning_rate": 1e-05,
+      "loss": 1.158932590484619,
+      "step": 4420
+    },
+    {
+      "epoch": 0.4663157894736842,
+      "grad_norm": 0.2987715900139601,
+      "learning_rate": 1e-05,
+      "loss": 1.1594856262207032,
+      "step": 4430
+    },
+    {
+      "epoch": 0.4673684210526316,
+      "grad_norm": 0.2861298282549877,
+      "learning_rate": 1e-05,
+      "loss": 1.1632608413696288,
+      "step": 4440
+    },
+    {
+      "epoch": 0.46842105263157896,
+      "grad_norm": 0.2779046888878849,
+      "learning_rate": 1e-05,
+      "loss": 1.1690286636352538,
+      "step": 4450
+    },
+    {
+      "epoch": 0.4694736842105263,
+      "grad_norm": 0.2969776932605365,
+      "learning_rate": 1e-05,
+      "loss": 1.1707868576049805,
+      "step": 4460
+    },
+    {
+      "epoch": 0.4705263157894737,
+      "grad_norm": 0.30486200002833497,
+      "learning_rate": 1e-05,
+      "loss": 1.1880663871765136,
+      "step": 4470
+    },
+    {
+      "epoch": 0.47157894736842104,
+      "grad_norm": 0.35135747830003167,
+      "learning_rate": 1e-05,
+      "loss": 1.175730323791504,
+      "step": 4480
+    },
+    {
+      "epoch": 0.4726315789473684,
+      "grad_norm": 0.2790550820773884,
+      "learning_rate": 1e-05,
+      "loss": 1.164478874206543,
+      "step": 4490
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.3178867063594838,
+      "learning_rate": 1e-05,
+      "loss": 1.1738519668579102,
+      "step": 4500
+    },
+    {
+      "epoch": 0.4747368421052632,
+      "grad_norm": 0.3189954363586918,
+      "learning_rate": 1e-05,
+      "loss": 1.1956332206726075,
+      "step": 4510
+    },
+    {
+      "epoch": 0.47578947368421054,
+      "grad_norm": 0.3106147939474411,
+      "learning_rate": 1e-05,
+      "loss": 1.1826568603515626,
+      "step": 4520
+    },
+    {
+      "epoch": 0.4768421052631579,
+      "grad_norm": 0.29558163483227895,
+      "learning_rate": 1e-05,
+      "loss": 1.1558403968811035,
+      "step": 4530
+    },
+    {
+      "epoch": 0.47789473684210526,
+      "grad_norm": 0.3154970989702764,
+      "learning_rate": 1e-05,
+      "loss": 1.1875129699707032,
+      "step": 4540
+    },
+    {
+      "epoch": 0.4789473684210526,
+      "grad_norm": 0.28838141531532535,
+      "learning_rate": 1e-05,
+      "loss": 1.16351375579834,
+      "step": 4550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27601846778759187,
+      "learning_rate": 1e-05,
+      "loss": 1.157202911376953,
+      "step": 4560
+    },
+    {
+      "epoch": 0.48105263157894734,
+      "grad_norm": 0.36272645375198076,
+      "learning_rate": 1e-05,
+      "loss": 1.1872642517089844,
+      "step": 4570
+    },
+    {
+      "epoch": 0.48210526315789476,
+      "grad_norm": 0.2996019252266834,
+      "learning_rate": 1e-05,
+      "loss": 1.1536483764648438,
+      "step": 4580
+    },
+    {
+      "epoch": 0.4831578947368421,
+      "grad_norm": 0.27601639952304935,
+      "learning_rate": 1e-05,
+      "loss": 1.1143065452575684,
+      "step": 4590
+    },
+    {
+      "epoch": 0.4842105263157895,
+      "grad_norm": 0.30258201078031255,
+      "learning_rate": 1e-05,
+      "loss": 1.0991535186767578,
+      "step": 4600
+    },
+    {
+      "epoch": 0.48526315789473684,
+      "grad_norm": 0.42339709127089475,
+      "learning_rate": 1e-05,
+      "loss": 1.0807291984558105,
+      "step": 4610
+    },
+    {
+      "epoch": 0.4863157894736842,
+      "grad_norm": 0.30433422283193823,
+      "learning_rate": 1e-05,
+      "loss": 1.055069923400879,
+      "step": 4620
+    },
+    {
+      "epoch": 0.48736842105263156,
+      "grad_norm": 0.27252972279928833,
+      "learning_rate": 1e-05,
+      "loss": 1.0379457473754883,
+      "step": 4630
+    },
+    {
+      "epoch": 0.4884210526315789,
+      "grad_norm": 0.4099710403446738,
+      "learning_rate": 1e-05,
+      "loss": 1.0420660972595215,
+      "step": 4640
+    },
+    {
+      "epoch": 0.48947368421052634,
+      "grad_norm": 0.27015960197020317,
+      "learning_rate": 1e-05,
+      "loss": 0.9960123062133789,
+      "step": 4650
+    },
+    {
+      "epoch": 0.4905263157894737,
+      "grad_norm": 0.2630280211080946,
+      "learning_rate": 1e-05,
+      "loss": 1.0044785499572755,
+      "step": 4660
+    },
+    {
+      "epoch": 0.49157894736842106,
+      "grad_norm": 0.2598889787175758,
+      "learning_rate": 1e-05,
+      "loss": 1.0007448196411133,
+      "step": 4670
+    },
+    {
+      "epoch": 0.4926315789473684,
+      "grad_norm": 0.27298421178866056,
+      "learning_rate": 1e-05,
+      "loss": 0.9759774208068848,
+      "step": 4680
+    },
+    {
+      "epoch": 0.4936842105263158,
+      "grad_norm": 0.267399121130361,
+      "learning_rate": 1e-05,
+      "loss": 0.9601457595825196,
+      "step": 4690
+    },
+    {
+      "epoch": 0.49473684210526314,
+      "grad_norm": 0.2498591485457542,
+      "learning_rate": 1e-05,
+      "loss": 0.9764652252197266,
+      "step": 4700
+    },
+    {
+      "epoch": 0.4957894736842105,
+      "grad_norm": 0.258699030521272,
+      "learning_rate": 1e-05,
+      "loss": 0.9989935874938964,
+      "step": 4710
+    },
+    {
+      "epoch": 0.4968421052631579,
+      "grad_norm": 0.24866080627110376,
+      "learning_rate": 1e-05,
+      "loss": 0.9777190208435058,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4978947368421053,
+      "grad_norm": 0.25304794580753004,
+      "learning_rate": 1e-05,
+      "loss": 0.9143146514892578,
+      "step": 4730
+    },
+    {
+      "epoch": 0.49894736842105264,
+      "grad_norm": 0.26300235903881375,
+      "learning_rate": 1e-05,
+      "loss": 0.9406648635864258,
+      "step": 4740
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2584782598440772,
+      "learning_rate": 1e-05,
+      "loss": 0.9758111000061035,
+      "step": 4750
+    },
+    {
+      "epoch": 0.5010526315789474,
+      "grad_norm": 0.258110700151434,
+      "learning_rate": 1e-05,
+      "loss": 0.9492972373962403,
+      "step": 4760
+    },
+    {
+      "epoch": 0.5021052631578947,
+      "grad_norm": 0.24918820723317286,
+      "learning_rate": 1e-05,
+      "loss": 0.931629753112793,
+      "step": 4770
+    },
+    {
+      "epoch": 0.5031578947368421,
+      "grad_norm": 0.2578190863958562,
+      "learning_rate": 1e-05,
+      "loss": 0.9528488159179688,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5042105263157894,
+      "grad_norm": 0.2465852899883327,
+      "learning_rate": 1e-05,
+      "loss": 0.9450368881225586,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5052631578947369,
+      "grad_norm": 0.23330795959601525,
+      "learning_rate": 1e-05,
+      "loss": 0.9238520622253418,
+      "step": 4800
+    },
+    {
+      "epoch": 0.5063157894736842,
+      "grad_norm": 0.24801816614440067,
+      "learning_rate": 1e-05,
+      "loss": 0.9313531875610351,
+      "step": 4810
+    },
+    {
+      "epoch": 0.5073684210526316,
+      "grad_norm": 0.31429072184462853,
+      "learning_rate": 1e-05,
+      "loss": 0.9388879776000977,
+      "step": 4820
+    },
+    {
+      "epoch": 0.508421052631579,
+      "grad_norm": 0.2625863523139627,
+      "learning_rate": 1e-05,
+      "loss": 0.9240442276000976,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5094736842105263,
+      "grad_norm": 0.25277184277510933,
+      "learning_rate": 1e-05,
+      "loss": 0.9421684265136718,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5105263157894737,
+      "grad_norm": 0.25311826090689316,
+      "learning_rate": 1e-05,
+      "loss": 0.9557743072509766,
+      "step": 4850
+    },
+    {
+      "epoch": 0.511578947368421,
+      "grad_norm": 0.24731903264539032,
+      "learning_rate": 1e-05,
+      "loss": 0.9125814437866211,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5126315789473684,
+      "grad_norm": 0.25640737202278496,
+      "learning_rate": 1e-05,
+      "loss": 0.9382814407348633,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5136842105263157,
+      "grad_norm": 0.24944189474155745,
+      "learning_rate": 1e-05,
+      "loss": 0.9284782409667969,
+      "step": 4880
+    },
+    {
+      "epoch": 0.5147368421052632,
+      "grad_norm": 0.26260837615313853,
+      "learning_rate": 1e-05,
+      "loss": 0.947537612915039,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5157894736842106,
+      "grad_norm": 0.25327581863441256,
+      "learning_rate": 1e-05,
+      "loss": 0.9095687866210938,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5168421052631579,
+      "grad_norm": 0.24595978826849177,
+      "learning_rate": 1e-05,
+      "loss": 0.9247632026672363,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5178947368421053,
+      "grad_norm": 0.2825462277295971,
+      "learning_rate": 1e-05,
+      "loss": 0.9398979187011719,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5189473684210526,
+      "grad_norm": 0.2695426481562848,
+      "learning_rate": 1e-05,
+      "loss": 0.9248984336853028,
+      "step": 4930
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3074442522069957,
+      "learning_rate": 1e-05,
+      "loss": 0.9281471252441407,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5210526315789473,
+      "grad_norm": 0.2516580543869498,
+      "learning_rate": 1e-05,
+      "loss": 0.9131274223327637,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5221052631578947,
+      "grad_norm": 0.27555495967470756,
+      "learning_rate": 1e-05,
+      "loss": 0.8868419647216796,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5231578947368422,
+      "grad_norm": 0.30466464049155256,
+      "learning_rate": 1e-05,
+      "loss": 0.915675163269043,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5242105263157895,
+      "grad_norm": 0.27594095303132987,
+      "learning_rate": 1e-05,
+      "loss": 0.9053061485290528,
+      "step": 4980
+    },
+    {
+      "epoch": 0.5252631578947369,
+      "grad_norm": 0.24209031636535625,
+      "learning_rate": 1e-05,
+      "loss": 0.8988632202148438,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.23233351503010502,
+      "learning_rate": 1e-05,
+      "loss": 0.8636146545410156,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5273684210526316,
+      "grad_norm": 0.25756110712647207,
+      "learning_rate": 1e-05,
+      "loss": 0.8804872512817383,
+      "step": 5010
+    },
+    {
+      "epoch": 0.5284210526315789,
+      "grad_norm": 0.23493102094534446,
+      "learning_rate": 1e-05,
+      "loss": 0.8646985054016113,
+      "step": 5020
+    },
+    {
+      "epoch": 0.5294736842105263,
+      "grad_norm": 0.9190229593813792,
+      "learning_rate": 1e-05,
+      "loss": 0.8608156204223633,
+      "step": 5030
+    },
+    {
+      "epoch": 0.5305263157894737,
+      "grad_norm": 0.23517783935433403,
+      "learning_rate": 1e-05,
+      "loss": 0.8259272575378418,
+      "step": 5040
+    },
+    {
+      "epoch": 0.531578947368421,
+      "grad_norm": 0.22708925781768755,
+      "learning_rate": 1e-05,
+      "loss": 0.8326313018798828,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5326315789473685,
+      "grad_norm": 0.29655823772902357,
+      "learning_rate": 1e-05,
+      "loss": 0.8479031562805176,
+      "step": 5060
+    },
+    {
+      "epoch": 0.5336842105263158,
+      "grad_norm": 0.29581499360363955,
+      "learning_rate": 1e-05,
+      "loss": 0.8843194961547851,
+      "step": 5070
+    },
+    {
+      "epoch": 0.5347368421052632,
+      "grad_norm": 0.2681316383303066,
+      "learning_rate": 1e-05,
+      "loss": 0.8562471389770507,
+      "step": 5080
+    },
+    {
+      "epoch": 0.5357894736842105,
+      "grad_norm": 0.7661874908592904,
+      "learning_rate": 1e-05,
+      "loss": 0.8677413940429688,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5368421052631579,
+      "grad_norm": 0.2765581758199201,
+      "learning_rate": 1e-05,
+      "loss": 0.9065875053405762,
+      "step": 5100
+    },
+    {
+      "epoch": 0.5378947368421053,
+      "grad_norm": 0.30942810112854024,
+      "learning_rate": 1e-05,
+      "loss": 0.8898434638977051,
+      "step": 5110
+    },
+    {
+      "epoch": 0.5389473684210526,
+      "grad_norm": 0.32284999083568194,
+      "learning_rate": 1e-05,
+      "loss": 0.9422303199768066,
+      "step": 5120
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29116737343846094,
+      "learning_rate": 1e-05,
+      "loss": 0.8884141921997071,
+      "step": 5130
+    },
+    {
+      "epoch": 0.5410526315789473,
+      "grad_norm": 0.26673236291585184,
+      "learning_rate": 1e-05,
+      "loss": 0.8889121055603028,
+      "step": 5140
+    },
+    {
+      "epoch": 0.5421052631578948,
+      "grad_norm": 0.2602052765972978,
+      "learning_rate": 1e-05,
+      "loss": 0.8402705192565918,
+      "step": 5150
+    },
+    {
+      "epoch": 0.5431578947368421,
+      "grad_norm": 0.5751678611132971,
+      "learning_rate": 1e-05,
+      "loss": 0.844548511505127,
+      "step": 5160
+    },
+    {
+      "epoch": 0.5442105263157895,
+      "grad_norm": 0.24620371771784214,
+      "learning_rate": 1e-05,
+      "loss": 0.854130744934082,
+      "step": 5170
+    },
+    {
+      "epoch": 0.5452631578947369,
+      "grad_norm": 0.24237745133805805,
+      "learning_rate": 1e-05,
+      "loss": 0.9062323570251465,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5463157894736842,
+      "grad_norm": 0.26770105449314396,
+      "learning_rate": 1e-05,
+      "loss": 0.85721435546875,
+      "step": 5190
+    },
+    {
+      "epoch": 0.5473684210526316,
+      "grad_norm": 0.24909991645571097,
+      "learning_rate": 1e-05,
+      "loss": 0.8664381980895997,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5484210526315789,
+      "grad_norm": 0.23810921919369427,
+      "learning_rate": 1e-05,
+      "loss": 0.8454217910766602,
+      "step": 5210
+    },
+    {
+      "epoch": 0.5494736842105263,
+      "grad_norm": 0.24364477284956954,
+      "learning_rate": 1e-05,
+      "loss": 0.8691944122314453,
+      "step": 5220
+    },
+    {
+      "epoch": 0.5505263157894736,
+      "grad_norm": 0.24996016515179256,
+      "learning_rate": 1e-05,
+      "loss": 0.8663163185119629,
+      "step": 5230
+    },
+    {
+      "epoch": 0.5515789473684211,
+      "grad_norm": 0.23728419202618237,
+      "learning_rate": 1e-05,
+      "loss": 0.8590461730957031,
+      "step": 5240
+    },
+    {
+      "epoch": 0.5526315789473685,
+      "grad_norm": 0.2393196911444887,
+      "learning_rate": 1e-05,
+      "loss": 0.8525073051452636,
+      "step": 5250
+    },
+    {
+      "epoch": 0.5536842105263158,
+      "grad_norm": 0.24100857985871957,
+      "learning_rate": 1e-05,
+      "loss": 0.8630128860473633,
+      "step": 5260
+    },
+    {
+      "epoch": 0.5547368421052632,
+      "grad_norm": 0.23645894070302922,
+      "learning_rate": 1e-05,
+      "loss": 0.9151544570922852,
+      "step": 5270
+    },
+    {
+      "epoch": 0.5557894736842105,
+      "grad_norm": 0.24537511624370215,
+      "learning_rate": 1e-05,
+      "loss": 0.9543111801147461,
+      "step": 5280
+    },
+    {
+      "epoch": 0.5568421052631579,
+      "grad_norm": 0.2576085202451571,
+      "learning_rate": 1e-05,
+      "loss": 0.9822376251220704,
+      "step": 5290
+    },
+    {
+      "epoch": 0.5578947368421052,
+      "grad_norm": 0.3117323397020148,
+      "learning_rate": 1e-05,
+      "loss": 1.00231876373291,
+      "step": 5300
+    },
+    {
+      "epoch": 0.5589473684210526,
+      "grad_norm": 0.26604246596351855,
+      "learning_rate": 1e-05,
+      "loss": 0.9745959281921387,
+      "step": 5310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2903626719487882,
+      "learning_rate": 1e-05,
+      "loss": 1.0009227752685548,
+      "step": 5320
+    },
+    {
+      "epoch": 0.5610526315789474,
+      "grad_norm": 0.2940872226490279,
+      "learning_rate": 1e-05,
+      "loss": 0.9908008575439453,
+      "step": 5330
+    },
+    {
+      "epoch": 0.5621052631578948,
+      "grad_norm": 0.29257244309343267,
+      "learning_rate": 1e-05,
+      "loss": 1.0361680030822753,
+      "step": 5340
+    },
+    {
+      "epoch": 0.5631578947368421,
+      "grad_norm": 0.27265931157308126,
+      "learning_rate": 1e-05,
+      "loss": 1.033724021911621,
+      "step": 5350
+    },
+    {
+      "epoch": 0.5642105263157895,
+      "grad_norm": 0.2851121850587208,
+      "learning_rate": 1e-05,
+      "loss": 1.0665639877319335,
+      "step": 5360
+    },
+    {
+      "epoch": 0.5652631578947368,
+      "grad_norm": 0.277506756097937,
+      "learning_rate": 1e-05,
+      "loss": 1.0624160766601562,
+      "step": 5370
+    },
+    {
+      "epoch": 0.5663157894736842,
+      "grad_norm": 0.32933400349916647,
+      "learning_rate": 1e-05,
+      "loss": 1.0804841995239258,
+      "step": 5380
+    },
+    {
+      "epoch": 0.5673684210526316,
+      "grad_norm": 0.307649261930152,
+      "learning_rate": 1e-05,
+      "loss": 1.0232925415039062,
+      "step": 5390
+    },
+    {
+      "epoch": 0.5684210526315789,
+      "grad_norm": 0.32783817695869727,
+      "learning_rate": 1e-05,
+      "loss": 1.0526607513427735,
+      "step": 5400
+    },
+    {
+      "epoch": 0.5694736842105264,
+      "grad_norm": 0.24856780525576358,
+      "learning_rate": 1e-05,
+      "loss": 1.0417704582214355,
+      "step": 5410
+    },
+    {
+      "epoch": 0.5705263157894737,
+      "grad_norm": 0.29516063019468686,
+      "learning_rate": 1e-05,
+      "loss": 1.0485565185546875,
+      "step": 5420
+    },
+    {
+      "epoch": 0.5715789473684211,
+      "grad_norm": 0.27069719325552716,
+      "learning_rate": 1e-05,
+      "loss": 1.09641752243042,
+      "step": 5430
+    },
+    {
+      "epoch": 0.5726315789473684,
+      "grad_norm": 0.2582520204620218,
+      "learning_rate": 1e-05,
+      "loss": 1.0912270545959473,
+      "step": 5440
+    },
+    {
+      "epoch": 0.5736842105263158,
+      "grad_norm": 0.2614495116390991,
+      "learning_rate": 1e-05,
+      "loss": 1.0877411842346192,
+      "step": 5450
+    },
+    {
+      "epoch": 0.5747368421052632,
+      "grad_norm": 0.27405339048633004,
+      "learning_rate": 1e-05,
+      "loss": 1.0787829399108886,
+      "step": 5460
+    },
+    {
+      "epoch": 0.5757894736842105,
+      "grad_norm": 0.26769810052419296,
+      "learning_rate": 1e-05,
+      "loss": 1.0821168899536133,
+      "step": 5470
+    },
+    {
+      "epoch": 0.5768421052631579,
+      "grad_norm": 0.26063449027647173,
+      "learning_rate": 1e-05,
+      "loss": 1.0660191535949708,
+      "step": 5480
+    },
+    {
+      "epoch": 0.5778947368421052,
+      "grad_norm": 0.2792441441763367,
+      "learning_rate": 1e-05,
+      "loss": 1.0737732887268066,
+      "step": 5490
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.3182976342568761,
+      "learning_rate": 1e-05,
+      "loss": 1.0843471527099608,
+      "step": 5500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2766411804683454,
+      "learning_rate": 1e-05,
+      "loss": 1.082973098754883,
+      "step": 5510
+    },
+    {
+      "epoch": 0.5810526315789474,
+      "grad_norm": 0.28381684819864317,
+      "learning_rate": 1e-05,
+      "loss": 1.070871353149414,
+      "step": 5520
+    },
+    {
+      "epoch": 0.5821052631578948,
+      "grad_norm": 0.2599404991560167,
+      "learning_rate": 1e-05,
+      "loss": 1.070331859588623,
+      "step": 5530
+    },
+    {
+      "epoch": 0.5831578947368421,
+      "grad_norm": 0.29725563220996526,
+      "learning_rate": 1e-05,
+      "loss": 1.0703227043151855,
+      "step": 5540
+    },
+    {
+      "epoch": 0.5842105263157895,
+      "grad_norm": 0.26272009320452683,
+      "learning_rate": 1e-05,
+      "loss": 1.072371482849121,
+      "step": 5550
+    },
+    {
+      "epoch": 0.5852631578947368,
+      "grad_norm": 0.25505284037586046,
+      "learning_rate": 1e-05,
+      "loss": 1.0749868392944335,
+      "step": 5560
+    },
+    {
+      "epoch": 0.5863157894736842,
+      "grad_norm": 0.26210031278913143,
+      "learning_rate": 1e-05,
+      "loss": 1.0864994049072265,
+      "step": 5570
+    },
+    {
+      "epoch": 0.5873684210526315,
+      "grad_norm": 0.2773584452812182,
+      "learning_rate": 1e-05,
+      "loss": 1.091717529296875,
+      "step": 5580
+    },
+    {
+      "epoch": 0.588421052631579,
+      "grad_norm": 0.27863914651262606,
+      "learning_rate": 1e-05,
+      "loss": 1.0727325439453126,
+      "step": 5590
+    },
+    {
+      "epoch": 0.5894736842105263,
+      "grad_norm": 0.2911662093515509,
+      "learning_rate": 1e-05,
+      "loss": 1.0701934814453125,
+      "step": 5600
+    },
+    {
+      "epoch": 0.5905263157894737,
+      "grad_norm": 0.25777215867093384,
+      "learning_rate": 1e-05,
+      "loss": 1.0899513244628907,
+      "step": 5610
+    },
+    {
+      "epoch": 0.5915789473684211,
+      "grad_norm": 0.2562991896853783,
+      "learning_rate": 1e-05,
+      "loss": 1.0459590911865235,
+      "step": 5620
+    },
+    {
+      "epoch": 0.5926315789473684,
+      "grad_norm": 0.2723446534709526,
+      "learning_rate": 1e-05,
+      "loss": 1.0656709671020508,
+      "step": 5630
+    },
+    {
+      "epoch": 0.5936842105263158,
+      "grad_norm": 0.277977646082909,
+      "learning_rate": 1e-05,
+      "loss": 1.069857406616211,
+      "step": 5640
+    },
+    {
+      "epoch": 0.5947368421052631,
+      "grad_norm": 1.7139848138958937,
+      "learning_rate": 1e-05,
+      "loss": 1.064733123779297,
+      "step": 5650
+    },
+    {
+      "epoch": 0.5957894736842105,
+      "grad_norm": 0.28247027269447345,
+      "learning_rate": 1e-05,
+      "loss": 1.0782355308532714,
+      "step": 5660
+    },
+    {
+      "epoch": 0.5968421052631578,
+      "grad_norm": 0.27028268690754587,
+      "learning_rate": 1e-05,
+      "loss": 1.0666091918945313,
+      "step": 5670
+    },
+    {
+      "epoch": 0.5978947368421053,
+      "grad_norm": 0.26011954989390823,
+      "learning_rate": 1e-05,
+      "loss": 1.088020133972168,
+      "step": 5680
+    },
+    {
+      "epoch": 0.5989473684210527,
+      "grad_norm": 0.27124697093601347,
+      "learning_rate": 1e-05,
+      "loss": 1.0595073699951172,
+      "step": 5690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4245999499915541,
+      "learning_rate": 1e-05,
+      "loss": 1.0736034393310547,
+      "step": 5700
+    },
+    {
+      "epoch": 0.6010526315789474,
+      "grad_norm": 0.2666023691408582,
+      "learning_rate": 1e-05,
+      "loss": 1.0937231063842774,
+      "step": 5710
+    },
+    {
+      "epoch": 0.6021052631578947,
+      "grad_norm": 0.255470498704476,
+      "learning_rate": 1e-05,
+      "loss": 1.0723279953002929,
+      "step": 5720
+    },
+    {
+      "epoch": 0.6031578947368421,
+      "grad_norm": 0.29897184411141275,
+      "learning_rate": 1e-05,
+      "loss": 1.0776229858398438,
+      "step": 5730
+    },
+    {
+      "epoch": 0.6042105263157894,
+      "grad_norm": 0.27758100851888856,
+      "learning_rate": 1e-05,
+      "loss": 1.0834442138671876,
+      "step": 5740
+    },
+    {
+      "epoch": 0.6052631578947368,
+      "grad_norm": 0.2676027730026317,
+      "learning_rate": 1e-05,
+      "loss": 1.0636725425720215,
+      "step": 5750
+    },
+    {
+      "epoch": 0.6063157894736843,
+      "grad_norm": 0.2988594706972973,
+      "learning_rate": 1e-05,
+      "loss": 1.0914352416992188,
+      "step": 5760
+    },
+    {
+      "epoch": 0.6073684210526316,
+      "grad_norm": 0.2630965419393403,
+      "learning_rate": 1e-05,
+      "loss": 1.0638557434082032,
+      "step": 5770
+    },
+    {
+      "epoch": 0.608421052631579,
+      "grad_norm": 0.29646744067661107,
+      "learning_rate": 1e-05,
+      "loss": 1.0554386138916017,
+      "step": 5780
+    },
+    {
+      "epoch": 0.6094736842105263,
+      "grad_norm": 0.27057955554541235,
+      "learning_rate": 1e-05,
+      "loss": 1.0474662780761719,
+      "step": 5790
+    },
+    {
+      "epoch": 0.6105263157894737,
+      "grad_norm": 0.2647050659936456,
+      "learning_rate": 1e-05,
+      "loss": 1.0582935333251953,
+      "step": 5800
+    },
+    {
+      "epoch": 0.611578947368421,
+      "grad_norm": 0.3370988721856255,
+      "learning_rate": 1e-05,
+      "loss": 1.0906749725341798,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6126315789473684,
+      "grad_norm": 0.28045526486414823,
+      "learning_rate": 1e-05,
+      "loss": 1.0644496917724608,
+      "step": 5820
+    },
+    {
+      "epoch": 0.6136842105263158,
+      "grad_norm": 0.274448519284646,
+      "learning_rate": 1e-05,
+      "loss": 1.078162384033203,
+      "step": 5830
+    },
+    {
+      "epoch": 0.6147368421052631,
+      "grad_norm": 0.26043712046572476,
+      "learning_rate": 1e-05,
+      "loss": 1.0445226669311523,
+      "step": 5840
+    },
+    {
+      "epoch": 0.6157894736842106,
+      "grad_norm": 0.31714316950945054,
+      "learning_rate": 1e-05,
+      "loss": 1.0405641555786134,
+      "step": 5850
+    },
+    {
+      "epoch": 0.6168421052631579,
+      "grad_norm": 0.2659167007536837,
+      "learning_rate": 1e-05,
+      "loss": 1.001264476776123,
+      "step": 5860
+    },
+    {
+      "epoch": 0.6178947368421053,
+      "grad_norm": 0.25855415657956754,
+      "learning_rate": 1e-05,
+      "loss": 0.9893880844116211,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6189473684210526,
+      "grad_norm": 0.2833524430866962,
+      "learning_rate": 1e-05,
+      "loss": 0.9901641845703125,
+      "step": 5880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.24746134749484275,
+      "learning_rate": 1e-05,
+      "loss": 0.9767881393432617,
+      "step": 5890
+    },
+    {
+      "epoch": 0.6210526315789474,
+      "grad_norm": 0.2490341392635219,
+      "learning_rate": 1e-05,
+      "loss": 0.9760961532592773,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6221052631578947,
+      "grad_norm": 0.24528379585798887,
+      "learning_rate": 1e-05,
+      "loss": 0.9478036880493164,
+      "step": 5910
+    },
+    {
+      "epoch": 0.6231578947368421,
+      "grad_norm": 0.27868509054807417,
+      "learning_rate": 1e-05,
+      "loss": 0.9562164306640625,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6242105263157894,
+      "grad_norm": 0.23838916093693996,
+      "learning_rate": 1e-05,
+      "loss": 0.9094364166259765,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6252631578947369,
+      "grad_norm": 0.22924112334759525,
+      "learning_rate": 1e-05,
+      "loss": 0.8979190826416016,
+      "step": 5940
+    },
+    {
+      "epoch": 0.6263157894736842,
+      "grad_norm": 0.24205525644029263,
+      "learning_rate": 1e-05,
+      "loss": 0.9365707397460937,
+      "step": 5950
+    },
+    {
+      "epoch": 0.6273684210526316,
+      "grad_norm": 0.24850922502286976,
+      "learning_rate": 1e-05,
+      "loss": 0.915318489074707,
+      "step": 5960
+    },
+    {
+      "epoch": 0.628421052631579,
+      "grad_norm": 0.22121831037855502,
+      "learning_rate": 1e-05,
+      "loss": 0.9018635749816895,
+      "step": 5970
+    },
+    {
+      "epoch": 0.6294736842105263,
+      "grad_norm": 1.1483261207152302,
+      "learning_rate": 1e-05,
+      "loss": 0.9099991798400879,
+      "step": 5980
+    },
+    {
+      "epoch": 0.6305263157894737,
+      "grad_norm": 0.2410279790982237,
+      "learning_rate": 1e-05,
+      "loss": 0.9137231826782226,
+      "step": 5990
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.2375306908477141,
+      "learning_rate": 1e-05,
+      "loss": 0.8785249710083007,
+      "step": 6000
+    },
+    {
+      "epoch": 0.6326315789473684,
+      "grad_norm": 0.23866325237067318,
+      "learning_rate": 1e-05,
+      "loss": 0.896181583404541,
+      "step": 6010
+    },
+    {
+      "epoch": 0.6336842105263157,
+      "grad_norm": 0.23208412343002432,
+      "learning_rate": 1e-05,
+      "loss": 0.874976634979248,
+      "step": 6020
+    },
+    {
+      "epoch": 0.6347368421052632,
+      "grad_norm": 0.2556461898863489,
+      "learning_rate": 1e-05,
+      "loss": 0.8863260269165039,
+      "step": 6030
+    },
+    {
+      "epoch": 0.6357894736842106,
+      "grad_norm": 0.24972578574988782,
+      "learning_rate": 1e-05,
+      "loss": 0.8927456855773925,
+      "step": 6040
+    },
+    {
+      "epoch": 0.6368421052631579,
+      "grad_norm": 0.23581160162060766,
+      "learning_rate": 1e-05,
+      "loss": 0.8682395935058593,
+      "step": 6050
+    },
+    {
+      "epoch": 0.6378947368421053,
+      "grad_norm": 0.23094574940093543,
+      "learning_rate": 1e-05,
+      "loss": 0.890745735168457,
+      "step": 6060
+    },
+    {
+      "epoch": 0.6389473684210526,
+      "grad_norm": 0.2230039802228098,
+      "learning_rate": 1e-05,
+      "loss": 0.8683330535888671,
+      "step": 6070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2540082631979619,
+      "learning_rate": 1e-05,
+      "loss": 0.8684318542480469,
+      "step": 6080
+    },
+    {
+      "epoch": 0.6410526315789473,
+      "grad_norm": 0.229321152076617,
+      "learning_rate": 1e-05,
+      "loss": 0.8784156799316406,
+      "step": 6090
+    },
+    {
+      "epoch": 0.6421052631578947,
+      "grad_norm": 0.2304400850675648,
+      "learning_rate": 1e-05,
+      "loss": 0.8877914428710938,
+      "step": 6100
+    },
+    {
+      "epoch": 0.6431578947368422,
+      "grad_norm": 0.3032858280685947,
+      "learning_rate": 1e-05,
+      "loss": 0.8660752296447753,
+      "step": 6110
+    },
+    {
+      "epoch": 0.6442105263157895,
+      "grad_norm": 0.23043843867769595,
+      "learning_rate": 1e-05,
+      "loss": 0.9108792304992676,
+      "step": 6120
+    },
+    {
+      "epoch": 0.6452631578947369,
+      "grad_norm": 0.2780232064537777,
+      "learning_rate": 1e-05,
+      "loss": 0.8758102416992187,
+      "step": 6130
+    },
+    {
+      "epoch": 0.6463157894736842,
+      "grad_norm": 0.22361969804560036,
+      "learning_rate": 1e-05,
+      "loss": 0.8758262634277344,
+      "step": 6140
+    },
+    {
+      "epoch": 0.6473684210526316,
+      "grad_norm": 0.22954502610636374,
+      "learning_rate": 1e-05,
+      "loss": 0.855862808227539,
+      "step": 6150
+    },
+    {
+      "epoch": 0.6484210526315789,
+      "grad_norm": 0.22568919674317806,
+      "learning_rate": 1e-05,
+      "loss": 0.8422542572021484,
+      "step": 6160
+    },
+    {
+      "epoch": 0.6494736842105263,
+      "grad_norm": 0.21095374288976398,
+      "learning_rate": 1e-05,
+      "loss": 0.8209218978881836,
+      "step": 6170
+    },
+    {
+      "epoch": 0.6505263157894737,
+      "grad_norm": 0.2380756495730225,
+      "learning_rate": 1e-05,
+      "loss": 0.8537910461425782,
+      "step": 6180
+    },
+    {
+      "epoch": 0.651578947368421,
+      "grad_norm": 0.23447822573827515,
+      "learning_rate": 1e-05,
+      "loss": 0.7929327487945557,
+      "step": 6190
+    },
+    {
+      "epoch": 0.6526315789473685,
+      "grad_norm": 0.2297877511968896,
+      "learning_rate": 1e-05,
+      "loss": 0.8187102317810059,
+      "step": 6200
+    },
+    {
+      "epoch": 0.6536842105263158,
+      "grad_norm": 0.23378270064613665,
+      "learning_rate": 1e-05,
+      "loss": 0.7985034942626953,
+      "step": 6210
+    },
+    {
+      "epoch": 0.6547368421052632,
+      "grad_norm": 0.2855349430667607,
+      "learning_rate": 1e-05,
+      "loss": 0.8041234970092773,
+      "step": 6220
+    },
+    {
+      "epoch": 0.6557894736842105,
+      "grad_norm": 0.24605610977528963,
+      "learning_rate": 1e-05,
+      "loss": 0.7719404220581054,
+      "step": 6230
+    },
+    {
+      "epoch": 0.6568421052631579,
+      "grad_norm": 0.2642172790860297,
+      "learning_rate": 1e-05,
+      "loss": 0.7462862491607666,
+      "step": 6240
+    },
+    {
+      "epoch": 0.6578947368421053,
+      "grad_norm": 0.27970912786467034,
+      "learning_rate": 1e-05,
+      "loss": 0.7801645278930665,
+      "step": 6250
+    },
+    {
+      "epoch": 0.6589473684210526,
+      "grad_norm": 0.2160390035966483,
+      "learning_rate": 1e-05,
+      "loss": 0.7723907470703125,
+      "step": 6260
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2172495851427012,
+      "learning_rate": 1e-05,
+      "loss": 0.7501685619354248,
+      "step": 6270
+    },
+    {
+      "epoch": 0.6610526315789473,
+      "grad_norm": 0.2197483200059199,
+      "learning_rate": 1e-05,
+      "loss": 0.7418710708618164,
+      "step": 6280
+    },
+    {
+      "epoch": 0.6621052631578948,
+      "grad_norm": 0.22562199127002897,
+      "learning_rate": 1e-05,
+      "loss": 0.7364078998565674,
+      "step": 6290
+    },
+    {
+      "epoch": 0.6631578947368421,
+      "grad_norm": 0.2487316717769606,
+      "learning_rate": 1e-05,
+      "loss": 0.756803321838379,
+      "step": 6300
+    },
+    {
+      "epoch": 0.6642105263157895,
+      "grad_norm": 0.21094326623762136,
+      "learning_rate": 1e-05,
+      "loss": 0.7114993095397949,
+      "step": 6310
+    },
+    {
+      "epoch": 0.6652631578947369,
+      "grad_norm": 0.5561025085003901,
+      "learning_rate": 1e-05,
+      "loss": 0.7145722389221192,
+      "step": 6320
+    },
+    {
+      "epoch": 0.6663157894736842,
+      "grad_norm": 0.2542687191246116,
+      "learning_rate": 1e-05,
+      "loss": 0.7226937294006348,
+      "step": 6330
+    },
+    {
+      "epoch": 0.6673684210526316,
+      "grad_norm": 0.2643466564198985,
+      "learning_rate": 1e-05,
+      "loss": 0.708247184753418,
+      "step": 6340
+    },
+    {
+      "epoch": 0.6684210526315789,
+      "grad_norm": 1.3709611857365271,
+      "learning_rate": 1e-05,
+      "loss": 0.7211179733276367,
+      "step": 6350
+    },
+    {
+      "epoch": 0.6694736842105263,
+      "grad_norm": 0.21994261788411282,
+      "learning_rate": 1e-05,
+      "loss": 0.7037035942077636,
+      "step": 6360
+    },
+    {
+      "epoch": 0.6705263157894736,
+      "grad_norm": 0.21226540625031654,
+      "learning_rate": 1e-05,
+      "loss": 0.713145399093628,
+      "step": 6370
+    },
+    {
+      "epoch": 0.671578947368421,
+      "grad_norm": 0.21208006446592162,
+      "learning_rate": 1e-05,
+      "loss": 0.7129218101501464,
+      "step": 6380
+    },
+    {
+      "epoch": 0.6726315789473685,
+      "grad_norm": 0.21257154493647612,
+      "learning_rate": 1e-05,
+      "loss": 0.7035690307617187,
+      "step": 6390
+    },
+    {
+      "epoch": 0.6736842105263158,
+      "grad_norm": 0.27203823294174123,
+      "learning_rate": 1e-05,
+      "loss": 0.7339433670043946,
+      "step": 6400
+    },
+    {
+      "epoch": 0.6747368421052632,
+      "grad_norm": 0.21881754935511646,
+      "learning_rate": 1e-05,
+      "loss": 0.7054609298706055,
+      "step": 6410
+    },
+    {
+      "epoch": 0.6757894736842105,
+      "grad_norm": 0.2098750364306379,
+      "learning_rate": 1e-05,
+      "loss": 0.6955869674682618,
+      "step": 6420
+    },
+    {
+      "epoch": 0.6768421052631579,
+      "grad_norm": 0.28362587688108587,
+      "learning_rate": 1e-05,
+      "loss": 0.6993033409118652,
+      "step": 6430
+    },
+    {
+      "epoch": 0.6778947368421052,
+      "grad_norm": 0.20629210163636771,
+      "learning_rate": 1e-05,
+      "loss": 0.7000613689422608,
+      "step": 6440
+    },
+    {
+      "epoch": 0.6789473684210526,
+      "grad_norm": 0.21206296257419424,
+      "learning_rate": 1e-05,
+      "loss": 0.6968954563140869,
+      "step": 6450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2265116528928763,
+      "learning_rate": 1e-05,
+      "loss": 0.6957263469696044,
+      "step": 6460
+    },
+    {
+      "epoch": 0.6810526315789474,
+      "grad_norm": 0.23414420776155453,
+      "learning_rate": 1e-05,
+      "loss": 0.7271471977233886,
+      "step": 6470
+    },
+    {
+      "epoch": 0.6821052631578948,
+      "grad_norm": 0.22370721771502153,
+      "learning_rate": 1e-05,
+      "loss": 0.7076205253601074,
+      "step": 6480
+    },
+    {
+      "epoch": 0.6831578947368421,
+      "grad_norm": 0.22220433460693967,
+      "learning_rate": 1e-05,
+      "loss": 0.6928388118743897,
+      "step": 6490
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.2247958776922717,
+      "learning_rate": 1e-05,
+      "loss": 0.6913023471832276,
+      "step": 6500
+    },
+    {
+      "epoch": 0.6852631578947368,
+      "grad_norm": 0.29001385417837844,
+      "learning_rate": 1e-05,
+      "loss": 0.6969108581542969,
+      "step": 6510
+    },
+    {
+      "epoch": 0.6863157894736842,
+      "grad_norm": 0.23179430205308796,
+      "learning_rate": 1e-05,
+      "loss": 0.6993899345397949,
+      "step": 6520
+    },
+    {
+      "epoch": 0.6873684210526316,
+      "grad_norm": 0.7731432450948373,
+      "learning_rate": 1e-05,
+      "loss": 0.7204256057739258,
+      "step": 6530
+    },
+    {
+      "epoch": 0.6884210526315789,
+      "grad_norm": 0.21339955377111022,
+      "learning_rate": 1e-05,
+      "loss": 0.687618637084961,
+      "step": 6540
+    },
+    {
+      "epoch": 0.6894736842105263,
+      "grad_norm": 0.22096033189707767,
+      "learning_rate": 1e-05,
+      "loss": 0.6879070281982422,
+      "step": 6550
+    },
+    {
+      "epoch": 0.6905263157894737,
+      "grad_norm": 0.19943501355754625,
+      "learning_rate": 1e-05,
+      "loss": 0.6515387535095215,
+      "step": 6560
+    },
+    {
+      "epoch": 0.6915789473684211,
+      "grad_norm": 0.22922295954951116,
+      "learning_rate": 1e-05,
+      "loss": 0.6515051364898682,
+      "step": 6570
+    },
+    {
+      "epoch": 0.6926315789473684,
+      "grad_norm": 0.20615036998840566,
+      "learning_rate": 1e-05,
+      "loss": 0.6451548576354981,
+      "step": 6580
+    },
+    {
+      "epoch": 0.6936842105263158,
+      "grad_norm": 0.21963201684482547,
+      "learning_rate": 1e-05,
+      "loss": 0.6539057731628418,
+      "step": 6590
+    },
+    {
+      "epoch": 0.6947368421052632,
+      "grad_norm": 0.21315606341405616,
+      "learning_rate": 1e-05,
+      "loss": 0.6748494625091552,
+      "step": 6600
+    },
+    {
+      "epoch": 0.6957894736842105,
+      "grad_norm": 0.20416413858841612,
+      "learning_rate": 1e-05,
+      "loss": 0.7066478729248047,
+      "step": 6610
+    },
+    {
+      "epoch": 0.6968421052631579,
+      "grad_norm": 0.22462071341021236,
+      "learning_rate": 1e-05,
+      "loss": 0.7325336933135986,
+      "step": 6620
+    },
+    {
+      "epoch": 0.6978947368421052,
+      "grad_norm": 0.2580875545867262,
+      "learning_rate": 1e-05,
+      "loss": 0.737650203704834,
+      "step": 6630
+    },
+    {
+      "epoch": 0.6989473684210527,
+      "grad_norm": 0.2300868817551716,
+      "learning_rate": 1e-05,
+      "loss": 0.7469221115112304,
+      "step": 6640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.21893365546024673,
+      "learning_rate": 1e-05,
+      "loss": 0.7442987442016602,
+      "step": 6650
+    },
+    {
+      "epoch": 0.7010526315789474,
+      "grad_norm": 0.22024866288239542,
+      "learning_rate": 1e-05,
+      "loss": 0.7899494171142578,
+      "step": 6660
+    },
+    {
+      "epoch": 0.7021052631578948,
+      "grad_norm": 0.22313458690080107,
+      "learning_rate": 1e-05,
+      "loss": 0.7705913543701172,
+      "step": 6670
+    },
+    {
+      "epoch": 0.7031578947368421,
+      "grad_norm": 0.2306828102095736,
+      "learning_rate": 1e-05,
+      "loss": 0.8105263710021973,
+      "step": 6680
+    },
+    {
+      "epoch": 0.7042105263157895,
+      "grad_norm": 0.21094125117347637,
+      "learning_rate": 1e-05,
+      "loss": 0.7712714672088623,
+      "step": 6690
+    },
+    {
+      "epoch": 0.7052631578947368,
+      "grad_norm": 0.21656705130937373,
+      "learning_rate": 1e-05,
+      "loss": 0.7751852989196777,
+      "step": 6700
+    },
+    {
+      "epoch": 0.7063157894736842,
+      "grad_norm": 0.22629428841151886,
+      "learning_rate": 1e-05,
+      "loss": 0.7708349227905273,
+      "step": 6710
+    },
+    {
+      "epoch": 0.7073684210526315,
+      "grad_norm": 0.23922086314770882,
+      "learning_rate": 1e-05,
+      "loss": 0.7916366577148437,
+      "step": 6720
+    },
+    {
+      "epoch": 0.708421052631579,
+      "grad_norm": 0.21099182467011027,
+      "learning_rate": 1e-05,
+      "loss": 0.8441093444824219,
+      "step": 6730
+    },
+    {
+      "epoch": 0.7094736842105264,
+      "grad_norm": 0.33818171593902324,
+      "learning_rate": 1e-05,
+      "loss": 0.7975322723388671,
+      "step": 6740
+    },
+    {
+      "epoch": 0.7105263157894737,
+      "grad_norm": 0.24825156508038823,
+      "learning_rate": 1e-05,
+      "loss": 0.794069242477417,
+      "step": 6750
+    },
+    {
+      "epoch": 0.7115789473684211,
+      "grad_norm": 0.22055862696416811,
+      "learning_rate": 1e-05,
+      "loss": 0.7830334663391113,
+      "step": 6760
+    },
+    {
+      "epoch": 0.7126315789473684,
+      "grad_norm": 0.2203005847917053,
+      "learning_rate": 1e-05,
+      "loss": 0.7622671127319336,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7136842105263158,
+      "grad_norm": 0.2197014248305324,
+      "learning_rate": 1e-05,
+      "loss": 0.7883934497833252,
+      "step": 6780
+    },
+    {
+      "epoch": 0.7147368421052631,
+      "grad_norm": 0.24252503943092846,
+      "learning_rate": 1e-05,
+      "loss": 0.7938870429992676,
+      "step": 6790
+    },
+    {
+      "epoch": 0.7157894736842105,
+      "grad_norm": 0.24121970439502166,
+      "learning_rate": 1e-05,
+      "loss": 0.7784067630767822,
+      "step": 6800
+    },
+    {
+      "epoch": 0.716842105263158,
+      "grad_norm": 0.2784368674999808,
+      "learning_rate": 1e-05,
+      "loss": 0.7677475929260253,
+      "step": 6810
+    },
+    {
+      "epoch": 0.7178947368421053,
+      "grad_norm": 0.2359049271276901,
+      "learning_rate": 1e-05,
+      "loss": 0.8072632789611817,
+      "step": 6820
+    },
+    {
+      "epoch": 0.7189473684210527,
+      "grad_norm": 0.22437907627890186,
+      "learning_rate": 1e-05,
+      "loss": 0.7935402393341064,
+      "step": 6830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20668779229830808,
+      "learning_rate": 1e-05,
+      "loss": 0.769287395477295,
+      "step": 6840
+    },
+    {
+      "epoch": 0.7210526315789474,
+      "grad_norm": 0.22116196449426564,
+      "learning_rate": 1e-05,
+      "loss": 0.8054431915283203,
+      "step": 6850
+    },
+    {
+      "epoch": 0.7221052631578947,
+      "grad_norm": 0.21693350555362054,
+      "learning_rate": 1e-05,
+      "loss": 0.8130393981933594,
+      "step": 6860
+    },
+    {
+      "epoch": 0.7231578947368421,
+      "grad_norm": 0.2384479116346583,
+      "learning_rate": 1e-05,
+      "loss": 0.8026229858398437,
+      "step": 6870
+    },
+    {
+      "epoch": 0.7242105263157895,
+      "grad_norm": 0.2527276166121089,
+      "learning_rate": 1e-05,
+      "loss": 0.8079002380371094,
+      "step": 6880
+    },
+    {
+      "epoch": 0.7252631578947368,
+      "grad_norm": 0.21867201305064243,
+      "learning_rate": 1e-05,
+      "loss": 0.8054183959960938,
+      "step": 6890
+    },
+    {
+      "epoch": 0.7263157894736842,
+      "grad_norm": 0.22873734944838955,
+      "learning_rate": 1e-05,
+      "loss": 0.789058780670166,
+      "step": 6900
+    },
+    {
+      "epoch": 0.7273684210526316,
+      "grad_norm": 0.2308172048008413,
+      "learning_rate": 1e-05,
+      "loss": 0.8146169662475586,
+      "step": 6910
+    },
+    {
+      "epoch": 0.728421052631579,
+      "grad_norm": 0.24122268833197355,
+      "learning_rate": 1e-05,
+      "loss": 0.7922428131103516,
+      "step": 6920
+    },
+    {
+      "epoch": 0.7294736842105263,
+      "grad_norm": 0.2589278588327487,
+      "learning_rate": 1e-05,
+      "loss": 0.8116107940673828,
+      "step": 6930
+    },
+    {
+      "epoch": 0.7305263157894737,
+      "grad_norm": 0.2422414713647026,
+      "learning_rate": 1e-05,
+      "loss": 0.8106143951416016,
+      "step": 6940
+    },
+    {
+      "epoch": 0.7315789473684211,
+      "grad_norm": 0.42135107683109013,
+      "learning_rate": 1e-05,
+      "loss": 0.8050684928894043,
+      "step": 6950
+    },
+    {
+      "epoch": 0.7326315789473684,
+      "grad_norm": 0.32189355128022173,
+      "learning_rate": 1e-05,
+      "loss": 0.809669017791748,
+      "step": 6960
+    },
+    {
+      "epoch": 0.7336842105263158,
+      "grad_norm": 0.25030376985073344,
+      "learning_rate": 1e-05,
+      "loss": 0.8026675224304199,
+      "step": 6970
+    },
+    {
+      "epoch": 0.7347368421052631,
+      "grad_norm": 0.21941824360356257,
+      "learning_rate": 1e-05,
+      "loss": 0.790980339050293,
+      "step": 6980
+    },
+    {
+      "epoch": 0.7357894736842105,
+      "grad_norm": 0.21245859747027523,
+      "learning_rate": 1e-05,
+      "loss": 0.7931976318359375,
+      "step": 6990
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.23517988949713722,
+      "learning_rate": 1e-05,
+      "loss": 0.7974249839782714,
+      "step": 7000
+    },
+    {
+      "epoch": 0.7378947368421053,
+      "grad_norm": 0.222519286171215,
+      "learning_rate": 1e-05,
+      "loss": 0.8180305480957031,
+      "step": 7010
+    },
+    {
+      "epoch": 0.7389473684210527,
+      "grad_norm": 0.21495493669956192,
+      "learning_rate": 1e-05,
+      "loss": 0.8175864219665527,
+      "step": 7020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.21989365599734784,
+      "learning_rate": 1e-05,
+      "loss": 0.7908399105072021,
+      "step": 7030
+    },
+    {
+      "epoch": 0.7410526315789474,
+      "grad_norm": 0.2427544195465207,
+      "learning_rate": 1e-05,
+      "loss": 0.7952071189880371,
+      "step": 7040
+    },
+    {
+      "epoch": 0.7421052631578947,
+      "grad_norm": 0.22650477903999935,
+      "learning_rate": 1e-05,
+      "loss": 0.7994481086730957,
+      "step": 7050
+    },
+    {
+      "epoch": 0.7431578947368421,
+      "grad_norm": 0.21357441607676197,
+      "learning_rate": 1e-05,
+      "loss": 0.8069334030151367,
+      "step": 7060
+    },
+    {
+      "epoch": 0.7442105263157894,
+      "grad_norm": 0.2201140724109808,
+      "learning_rate": 1e-05,
+      "loss": 0.8093648910522461,
+      "step": 7070
+    },
+    {
+      "epoch": 0.7452631578947368,
+      "grad_norm": 0.24051186131749505,
+      "learning_rate": 1e-05,
+      "loss": 0.7903453350067139,
+      "step": 7080
+    },
+    {
+      "epoch": 0.7463157894736843,
+      "grad_norm": 0.22581776759825037,
+      "learning_rate": 1e-05,
+      "loss": 0.7947319030761719,
+      "step": 7090
+    },
+    {
+      "epoch": 0.7473684210526316,
+      "grad_norm": 0.22080360755104417,
+      "learning_rate": 1e-05,
+      "loss": 0.8143038749694824,
+      "step": 7100
+    },
+    {
+      "epoch": 0.748421052631579,
+      "grad_norm": 0.3771997592677159,
+      "learning_rate": 1e-05,
+      "loss": 0.7976189613342285,
+      "step": 7110
+    },
+    {
+      "epoch": 0.7494736842105263,
+      "grad_norm": 0.22213107241717228,
+      "learning_rate": 1e-05,
+      "loss": 0.8067026138305664,
+      "step": 7120
+    },
+    {
+      "epoch": 0.7505263157894737,
+      "grad_norm": 0.2718975257239632,
+      "learning_rate": 1e-05,
+      "loss": 0.8294075012207032,
+      "step": 7130
+    },
+    {
+      "epoch": 0.751578947368421,
+      "grad_norm": 0.23780411464333923,
+      "learning_rate": 1e-05,
+      "loss": 0.7874054908752441,
+      "step": 7140
+    },
+    {
+      "epoch": 0.7526315789473684,
+      "grad_norm": 0.23725747282111376,
+      "learning_rate": 1e-05,
+      "loss": 0.7973507881164551,
+      "step": 7150
+    },
+    {
+      "epoch": 0.7536842105263157,
+      "grad_norm": 0.24077045352814438,
+      "learning_rate": 1e-05,
+      "loss": 0.7958509922027588,
+      "step": 7160
+    },
+    {
+      "epoch": 0.7547368421052632,
+      "grad_norm": 0.2982923938167098,
+      "learning_rate": 1e-05,
+      "loss": 0.8160445213317871,
+      "step": 7170
+    },
+    {
+      "epoch": 0.7557894736842106,
+      "grad_norm": 0.23719355121322688,
+      "learning_rate": 1e-05,
+      "loss": 0.8023138999938965,
+      "step": 7180
+    },
+    {
+      "epoch": 0.7568421052631579,
+      "grad_norm": 0.24486210177890705,
+      "learning_rate": 1e-05,
+      "loss": 0.8119283676147461,
+      "step": 7190
+    },
+    {
+      "epoch": 0.7578947368421053,
+      "grad_norm": 0.21943228478242519,
+      "learning_rate": 1e-05,
+      "loss": 0.8197574615478516,
+      "step": 7200
+    },
+    {
+      "epoch": 0.7589473684210526,
+      "grad_norm": 0.22261923138934933,
+      "learning_rate": 1e-05,
+      "loss": 0.7940950393676758,
+      "step": 7210
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.23837799196255208,
+      "learning_rate": 1e-05,
+      "loss": 0.8647306442260743,
+      "step": 7220
+    },
+    {
+      "epoch": 0.7610526315789473,
+      "grad_norm": 0.23995977027261564,
+      "learning_rate": 1e-05,
+      "loss": 0.87930908203125,
+      "step": 7230
+    },
+    {
+      "epoch": 0.7621052631578947,
+      "grad_norm": 0.2480329560166419,
+      "learning_rate": 1e-05,
+      "loss": 0.8913862228393554,
+      "step": 7240
+    },
+    {
+      "epoch": 0.7631578947368421,
+      "grad_norm": 0.24721898244174687,
+      "learning_rate": 1e-05,
+      "loss": 0.9122695922851562,
+      "step": 7250
+    },
+    {
+      "epoch": 0.7642105263157895,
+      "grad_norm": 0.32129027032976415,
+      "learning_rate": 1e-05,
+      "loss": 0.8943394660949707,
+      "step": 7260
+    },
+    {
+      "epoch": 0.7652631578947369,
+      "grad_norm": 0.387445678252169,
+      "learning_rate": 1e-05,
+      "loss": 0.921629524230957,
+      "step": 7270
+    },
+    {
+      "epoch": 0.7663157894736842,
+      "grad_norm": 0.31670043315980667,
+      "learning_rate": 1e-05,
+      "loss": 0.9350265502929688,
+      "step": 7280
+    },
+    {
+      "epoch": 0.7673684210526316,
+      "grad_norm": 0.2684366897375879,
+      "learning_rate": 1e-05,
+      "loss": 0.9515501022338867,
+      "step": 7290
+    },
+    {
+      "epoch": 0.7684210526315789,
+      "grad_norm": 0.22711858709116733,
+      "learning_rate": 1e-05,
+      "loss": 0.9535702705383301,
+      "step": 7300
+    },
+    {
+      "epoch": 0.7694736842105263,
+      "grad_norm": 0.22564934263447756,
+      "learning_rate": 1e-05,
+      "loss": 0.960813331604004,
+      "step": 7310
+    },
+    {
+      "epoch": 0.7705263157894737,
+      "grad_norm": 0.3029809294451466,
+      "learning_rate": 1e-05,
+      "loss": 0.9454540252685547,
+      "step": 7320
+    },
+    {
+      "epoch": 0.771578947368421,
+      "grad_norm": 0.24579654409210522,
+      "learning_rate": 1e-05,
+      "loss": 0.9875738143920898,
+      "step": 7330
+    },
+    {
+      "epoch": 0.7726315789473684,
+      "grad_norm": 0.24318621684305228,
+      "learning_rate": 1e-05,
+      "loss": 0.9857606887817383,
+      "step": 7340
+    },
+    {
+      "epoch": 0.7736842105263158,
+      "grad_norm": 0.2539040801315566,
+      "learning_rate": 1e-05,
+      "loss": 0.9809976577758789,
+      "step": 7350
+    },
+    {
+      "epoch": 0.7747368421052632,
+      "grad_norm": 0.2740015820976985,
+      "learning_rate": 1e-05,
+      "loss": 0.9719850540161132,
+      "step": 7360
+    },
+    {
+      "epoch": 0.7757894736842105,
+      "grad_norm": 0.2378094064983973,
+      "learning_rate": 1e-05,
+      "loss": 0.9772319793701172,
+      "step": 7370
+    },
+    {
+      "epoch": 0.7768421052631579,
+      "grad_norm": 0.27705766171370405,
+      "learning_rate": 1e-05,
+      "loss": 0.9895281791687012,
+      "step": 7380
+    },
+    {
+      "epoch": 0.7778947368421053,
+      "grad_norm": 0.24196561701464475,
+      "learning_rate": 1e-05,
+      "loss": 0.9818606376647949,
+      "step": 7390
+    },
+    {
+      "epoch": 0.7789473684210526,
+      "grad_norm": 0.25970230004728706,
+      "learning_rate": 1e-05,
+      "loss": 1.0087711334228515,
+      "step": 7400
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.27070734077316283,
+      "learning_rate": 1e-05,
+      "loss": 0.9816142082214355,
+      "step": 7410
+    },
+    {
+      "epoch": 0.7810526315789473,
+      "grad_norm": 0.2844462100920115,
+      "learning_rate": 1e-05,
+      "loss": 1.0093915939331055,
+      "step": 7420
+    },
+    {
+      "epoch": 0.7821052631578947,
+      "grad_norm": 0.2594366916260847,
+      "learning_rate": 1e-05,
+      "loss": 0.9957893371582032,
+      "step": 7430
+    },
+    {
+      "epoch": 0.783157894736842,
+      "grad_norm": 0.24648171036866237,
+      "learning_rate": 1e-05,
+      "loss": 1.0256897926330566,
+      "step": 7440
+    },
+    {
+      "epoch": 0.7842105263157895,
+      "grad_norm": 0.3326912831027331,
+      "learning_rate": 1e-05,
+      "loss": 1.0094666481018066,
+      "step": 7450
+    },
+    {
+      "epoch": 0.7852631578947369,
+      "grad_norm": 0.3334474798680566,
+      "learning_rate": 1e-05,
+      "loss": 1.0086650848388672,
+      "step": 7460
+    },
+    {
+      "epoch": 0.7863157894736842,
+      "grad_norm": 0.30131576125073195,
+      "learning_rate": 1e-05,
+      "loss": 1.0258222579956056,
+      "step": 7470
+    },
+    {
+      "epoch": 0.7873684210526316,
+      "grad_norm": 0.23183581804350187,
+      "learning_rate": 1e-05,
+      "loss": 1.0068394660949707,
+      "step": 7480
+    },
+    {
+      "epoch": 0.7884210526315789,
+      "grad_norm": 0.2730390358942403,
+      "learning_rate": 1e-05,
+      "loss": 0.9967912673950196,
+      "step": 7490
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.23976652211837599,
+      "learning_rate": 1e-05,
+      "loss": 1.0071218490600586,
+      "step": 7500
+    },
+    {
+      "epoch": 0.7905263157894736,
+      "grad_norm": 0.24687748192505013,
+      "learning_rate": 1e-05,
+      "loss": 1.004594898223877,
+      "step": 7510
+    },
+    {
+      "epoch": 0.791578947368421,
+      "grad_norm": 0.24573823935970499,
+      "learning_rate": 1e-05,
+      "loss": 1.0208873748779297,
+      "step": 7520
+    },
+    {
+      "epoch": 0.7926315789473685,
+      "grad_norm": 0.2691649930769953,
+      "learning_rate": 1e-05,
+      "loss": 1.0248262405395507,
+      "step": 7530
+    },
+    {
+      "epoch": 0.7936842105263158,
+      "grad_norm": 0.4248621901862408,
+      "learning_rate": 1e-05,
+      "loss": 1.0332860946655273,
+      "step": 7540
+    },
+    {
+      "epoch": 0.7947368421052632,
+      "grad_norm": 0.2705049287313093,
+      "learning_rate": 1e-05,
+      "loss": 0.995975112915039,
+      "step": 7550
+    },
+    {
+      "epoch": 0.7957894736842105,
+      "grad_norm": 0.28653098115122744,
+      "learning_rate": 1e-05,
+      "loss": 1.0159503936767578,
+      "step": 7560
+    },
+    {
+      "epoch": 0.7968421052631579,
+      "grad_norm": 0.29178481913543475,
+      "learning_rate": 1e-05,
+      "loss": 1.0173696517944335,
+      "step": 7570
+    },
+    {
+      "epoch": 0.7978947368421052,
+      "grad_norm": 0.2734088082876844,
+      "learning_rate": 1e-05,
+      "loss": 1.0264354705810548,
+      "step": 7580
+    },
+    {
+      "epoch": 0.7989473684210526,
+      "grad_norm": 0.261492476261997,
+      "learning_rate": 1e-05,
+      "loss": 1.0321282386779784,
+      "step": 7590
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2470753720478336,
+      "learning_rate": 1e-05,
+      "loss": 1.0191498756408692,
+      "step": 7600
+    },
+    {
+      "epoch": 0.8010526315789473,
+      "grad_norm": 0.24059955821517787,
+      "learning_rate": 1e-05,
+      "loss": 1.014815902709961,
+      "step": 7610
+    },
+    {
+      "epoch": 0.8021052631578948,
+      "grad_norm": 0.2493025233644774,
+      "learning_rate": 1e-05,
+      "loss": 1.0249458312988282,
+      "step": 7620
+    },
+    {
+      "epoch": 0.8031578947368421,
+      "grad_norm": 0.25193040573294745,
+      "learning_rate": 1e-05,
+      "loss": 1.0201826095581055,
+      "step": 7630
+    },
+    {
+      "epoch": 0.8042105263157895,
+      "grad_norm": 0.34953487134760375,
+      "learning_rate": 1e-05,
+      "loss": 1.02661190032959,
+      "step": 7640
+    },
+    {
+      "epoch": 0.8052631578947368,
+      "grad_norm": 0.3429544066348788,
+      "learning_rate": 1e-05,
+      "loss": 1.0370089530944824,
+      "step": 7650
+    },
+    {
+      "epoch": 0.8063157894736842,
+      "grad_norm": 0.26265155219421216,
+      "learning_rate": 1e-05,
+      "loss": 1.0190326690673828,
+      "step": 7660
+    },
+    {
+      "epoch": 0.8073684210526316,
+      "grad_norm": 0.2646612799609565,
+      "learning_rate": 1e-05,
+      "loss": 1.015769100189209,
+      "step": 7670
+    },
+    {
+      "epoch": 0.8084210526315789,
+      "grad_norm": 0.2770433943915144,
+      "learning_rate": 1e-05,
+      "loss": 1.0137809753417968,
+      "step": 7680
+    },
+    {
+      "epoch": 0.8094736842105263,
+      "grad_norm": 0.26086533194363315,
+      "learning_rate": 1e-05,
+      "loss": 1.0212547302246093,
+      "step": 7690
+    },
+    {
+      "epoch": 0.8105263157894737,
+      "grad_norm": 0.2645076293992011,
+      "learning_rate": 1e-05,
+      "loss": 1.0370710372924805,
+      "step": 7700
+    },
+    {
+      "epoch": 0.8115789473684211,
+      "grad_norm": 0.2696951338822986,
+      "learning_rate": 1e-05,
+      "loss": 1.0568577766418457,
+      "step": 7710
+    },
+    {
+      "epoch": 0.8126315789473684,
+      "grad_norm": 0.2642200813941431,
+      "learning_rate": 1e-05,
+      "loss": 1.0690131187438965,
+      "step": 7720
+    },
+    {
+      "epoch": 0.8136842105263158,
+      "grad_norm": 0.29336846930639393,
+      "learning_rate": 1e-05,
+      "loss": 1.1093889236450196,
+      "step": 7730
+    },
+    {
+      "epoch": 0.8147368421052632,
+      "grad_norm": 0.3747212784807069,
+      "learning_rate": 1e-05,
+      "loss": 1.1603777885437012,
+      "step": 7740
+    },
+    {
+      "epoch": 0.8157894736842105,
+      "grad_norm": 0.28430975732957897,
+      "learning_rate": 1e-05,
+      "loss": 1.163645362854004,
+      "step": 7750
+    },
+    {
+      "epoch": 0.8168421052631579,
+      "grad_norm": 0.27600699550474717,
+      "learning_rate": 1e-05,
+      "loss": 1.1624645233154296,
+      "step": 7760
+    },
+    {
+      "epoch": 0.8178947368421052,
+      "grad_norm": 0.4433050757013603,
+      "learning_rate": 1e-05,
+      "loss": 1.183049774169922,
+      "step": 7770
+    },
+    {
+      "epoch": 0.8189473684210526,
+      "grad_norm": 0.273714854354326,
+      "learning_rate": 1e-05,
+      "loss": 1.1642326354980468,
+      "step": 7780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27637324470956354,
+      "learning_rate": 1e-05,
+      "loss": 1.1787050247192383,
+      "step": 7790
+    },
+    {
+      "epoch": 0.8210526315789474,
+      "grad_norm": 0.2587899888920335,
+      "learning_rate": 1e-05,
+      "loss": 1.1967472076416015,
+      "step": 7800
+    },
+    {
+      "epoch": 0.8221052631578948,
+      "grad_norm": 0.2662067464157399,
+      "learning_rate": 1e-05,
+      "loss": 1.1967985153198242,
+      "step": 7810
+    },
+    {
+      "epoch": 0.8231578947368421,
+      "grad_norm": 0.2556056301094144,
+      "learning_rate": 1e-05,
+      "loss": 1.1737977981567382,
+      "step": 7820
+    },
+    {
+      "epoch": 0.8242105263157895,
+      "grad_norm": 0.38555554895987826,
+      "learning_rate": 1e-05,
+      "loss": 1.2106051445007324,
+      "step": 7830
+    },
+    {
+      "epoch": 0.8252631578947368,
+      "grad_norm": 0.26601130390001754,
+      "learning_rate": 1e-05,
+      "loss": 1.2188627243041992,
+      "step": 7840
+    },
+    {
+      "epoch": 0.8263157894736842,
+      "grad_norm": 0.29675333856574926,
+      "learning_rate": 1e-05,
+      "loss": 1.2107175827026366,
+      "step": 7850
+    },
+    {
+      "epoch": 0.8273684210526315,
+      "grad_norm": 0.28922328984457124,
+      "learning_rate": 1e-05,
+      "loss": 1.2110652923583984,
+      "step": 7860
+    },
+    {
+      "epoch": 0.828421052631579,
+      "grad_norm": 0.306666401156183,
+      "learning_rate": 1e-05,
+      "loss": 1.2091556549072267,
+      "step": 7870
+    },
+    {
+      "epoch": 0.8294736842105264,
+      "grad_norm": 0.2712439917601487,
+      "learning_rate": 1e-05,
+      "loss": 1.1936758041381836,
+      "step": 7880
+    },
+    {
+      "epoch": 0.8305263157894737,
+      "grad_norm": 0.31304490150174646,
+      "learning_rate": 1e-05,
+      "loss": 1.2132518768310547,
+      "step": 7890
+    },
+    {
+      "epoch": 0.8315789473684211,
+      "grad_norm": 0.2644723175451005,
+      "learning_rate": 1e-05,
+      "loss": 1.2193326950073242,
+      "step": 7900
+    },
+    {
+      "epoch": 0.8326315789473684,
+      "grad_norm": 0.26592616156636745,
+      "learning_rate": 1e-05,
+      "loss": 1.210757064819336,
+      "step": 7910
+    },
+    {
+      "epoch": 0.8336842105263158,
+      "grad_norm": 0.39029565900173574,
+      "learning_rate": 1e-05,
+      "loss": 1.225381851196289,
+      "step": 7920
+    },
+    {
+      "epoch": 0.8347368421052631,
+      "grad_norm": 0.30751609974574706,
+      "learning_rate": 1e-05,
+      "loss": 1.2244022369384766,
+      "step": 7930
+    },
+    {
+      "epoch": 0.8357894736842105,
+      "grad_norm": 0.2817083146221133,
+      "learning_rate": 1e-05,
+      "loss": 1.2566282272338867,
+      "step": 7940
+    },
+    {
+      "epoch": 0.8368421052631579,
+      "grad_norm": 0.26636277276418757,
+      "learning_rate": 1e-05,
+      "loss": 1.229690933227539,
+      "step": 7950
+    },
+    {
+      "epoch": 0.8378947368421052,
+      "grad_norm": 0.288134382279876,
+      "learning_rate": 1e-05,
+      "loss": 1.2371486663818358,
+      "step": 7960
+    },
+    {
+      "epoch": 0.8389473684210527,
+      "grad_norm": 0.2896289081751357,
+      "learning_rate": 1e-05,
+      "loss": 1.2217337608337402,
+      "step": 7970
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2713476353430279,
+      "learning_rate": 1e-05,
+      "loss": 1.2283599853515625,
+      "step": 7980
+    },
+    {
+      "epoch": 0.8410526315789474,
+      "grad_norm": 0.28393438845155,
+      "learning_rate": 1e-05,
+      "loss": 1.2311813354492187,
+      "step": 7990
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.3205106371080702,
+      "learning_rate": 1e-05,
+      "loss": 1.2315404891967774,
+      "step": 8000
+    },
+    {
+      "epoch": 0.8431578947368421,
+      "grad_norm": 0.271140189790355,
+      "learning_rate": 1e-05,
+      "loss": 1.226878547668457,
+      "step": 8010
+    },
+    {
+      "epoch": 0.8442105263157895,
+      "grad_norm": 0.2768312930367996,
+      "learning_rate": 1e-05,
+      "loss": 1.233083152770996,
+      "step": 8020
+    },
+    {
+      "epoch": 0.8452631578947368,
+      "grad_norm": 0.312013236093156,
+      "learning_rate": 1e-05,
+      "loss": 1.2304747581481934,
+      "step": 8030
+    },
+    {
+      "epoch": 0.8463157894736842,
+      "grad_norm": 0.262525314868118,
+      "learning_rate": 1e-05,
+      "loss": 1.2518205642700195,
+      "step": 8040
+    },
+    {
+      "epoch": 0.8473684210526315,
+      "grad_norm": 0.3137428501274644,
+      "learning_rate": 1e-05,
+      "loss": 1.230078125,
+      "step": 8050
+    },
+    {
+      "epoch": 0.848421052631579,
+      "grad_norm": 0.2884688333055249,
+      "learning_rate": 1e-05,
+      "loss": 1.2260973930358887,
+      "step": 8060
+    },
+    {
+      "epoch": 0.8494736842105263,
+      "grad_norm": 0.30946310048072445,
+      "learning_rate": 1e-05,
+      "loss": 1.2276168823242188,
+      "step": 8070
+    },
+    {
+      "epoch": 0.8505263157894737,
+      "grad_norm": 0.2674595172289079,
+      "learning_rate": 1e-05,
+      "loss": 1.2486080169677733,
+      "step": 8080
+    },
+    {
+      "epoch": 0.8515789473684211,
+      "grad_norm": 0.2997546586811151,
+      "learning_rate": 1e-05,
+      "loss": 1.2351938247680665,
+      "step": 8090
+    },
+    {
+      "epoch": 0.8526315789473684,
+      "grad_norm": 0.3049641872347451,
+      "learning_rate": 1e-05,
+      "loss": 1.237974262237549,
+      "step": 8100
+    },
+    {
+      "epoch": 0.8536842105263158,
+      "grad_norm": 0.2744867853610244,
+      "learning_rate": 1e-05,
+      "loss": 1.2470075607299804,
+      "step": 8110
+    },
+    {
+      "epoch": 0.8547368421052631,
+      "grad_norm": 0.2792616538196227,
+      "learning_rate": 1e-05,
+      "loss": 1.2331491470336915,
+      "step": 8120
+    },
+    {
+      "epoch": 0.8557894736842105,
+      "grad_norm": 0.27100610937087705,
+      "learning_rate": 1e-05,
+      "loss": 1.2516610145568847,
+      "step": 8130
+    },
+    {
+      "epoch": 0.8568421052631578,
+      "grad_norm": 0.2945693112909019,
+      "learning_rate": 1e-05,
+      "loss": 1.2330408096313477,
+      "step": 8140
+    },
+    {
+      "epoch": 0.8578947368421053,
+      "grad_norm": 0.3191745059450007,
+      "learning_rate": 1e-05,
+      "loss": 1.2524785041809081,
+      "step": 8150
+    },
+    {
+      "epoch": 0.8589473684210527,
+      "grad_norm": 0.322160297237838,
+      "learning_rate": 1e-05,
+      "loss": 1.2457155227661132,
+      "step": 8160
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2982507984823821,
+      "learning_rate": 1e-05,
+      "loss": 1.2377269744873047,
+      "step": 8170
+    },
+    {
+      "epoch": 0.8610526315789474,
+      "grad_norm": 0.2993102434394868,
+      "learning_rate": 1e-05,
+      "loss": 1.2454538345336914,
+      "step": 8180
+    },
+    {
+      "epoch": 0.8621052631578947,
+      "grad_norm": 0.2667807937841803,
+      "learning_rate": 1e-05,
+      "loss": 1.2359396934509277,
+      "step": 8190
+    },
+    {
+      "epoch": 0.8631578947368421,
+      "grad_norm": 0.28120038112128,
+      "learning_rate": 1e-05,
+      "loss": 1.2541059494018554,
+      "step": 8200
+    },
+    {
+      "epoch": 0.8642105263157894,
+      "grad_norm": 0.25963944675083334,
+      "learning_rate": 1e-05,
+      "loss": 1.2363154411315918,
+      "step": 8210
+    },
+    {
+      "epoch": 0.8652631578947368,
+      "grad_norm": 0.29092482264622005,
+      "learning_rate": 1e-05,
+      "loss": 1.2436670303344726,
+      "step": 8220
+    },
+    {
+      "epoch": 0.8663157894736843,
+      "grad_norm": 0.26291089899810094,
+      "learning_rate": 1e-05,
+      "loss": 1.2464160919189453,
+      "step": 8230
+    },
+    {
+      "epoch": 0.8673684210526316,
+      "grad_norm": 0.34922903820150786,
+      "learning_rate": 1e-05,
+      "loss": 1.2590576171875,
+      "step": 8240
+    },
+    {
+      "epoch": 0.868421052631579,
+      "grad_norm": 0.30060901133395024,
+      "learning_rate": 1e-05,
+      "loss": 1.2264986991882325,
+      "step": 8250
+    },
+    {
+      "epoch": 0.8694736842105263,
+      "grad_norm": 0.29273411855631515,
+      "learning_rate": 1e-05,
+      "loss": 1.2725407600402832,
+      "step": 8260
+    },
+    {
+      "epoch": 0.8705263157894737,
+      "grad_norm": 0.27022805398583977,
+      "learning_rate": 1e-05,
+      "loss": 1.2435733795166015,
+      "step": 8270
+    },
+    {
+      "epoch": 0.871578947368421,
+      "grad_norm": 0.27936632971869374,
+      "learning_rate": 1e-05,
+      "loss": 1.2213335037231445,
+      "step": 8280
+    },
+    {
+      "epoch": 0.8726315789473684,
+      "grad_norm": 0.3288218836779532,
+      "learning_rate": 1e-05,
+      "loss": 1.2252776145935058,
+      "step": 8290
+    },
+    {
+      "epoch": 0.8736842105263158,
+      "grad_norm": 0.2732197293819526,
+      "learning_rate": 1e-05,
+      "loss": 1.248112392425537,
+      "step": 8300
+    },
+    {
+      "epoch": 0.8747368421052631,
+      "grad_norm": 0.2567482594715607,
+      "learning_rate": 1e-05,
+      "loss": 1.2342848777770996,
+      "step": 8310
+    },
+    {
+      "epoch": 0.8757894736842106,
+      "grad_norm": 0.3031672698729003,
+      "learning_rate": 1e-05,
+      "loss": 1.2218093872070312,
+      "step": 8320
+    },
+    {
+      "epoch": 0.8768421052631579,
+      "grad_norm": 0.2736199897484509,
+      "learning_rate": 1e-05,
+      "loss": 1.2322706222534179,
+      "step": 8330
+    },
+    {
+      "epoch": 0.8778947368421053,
+      "grad_norm": 0.2896898314563102,
+      "learning_rate": 1e-05,
+      "loss": 1.2112766265869142,
+      "step": 8340
+    },
+    {
+      "epoch": 0.8789473684210526,
+      "grad_norm": 0.2854652985502235,
+      "learning_rate": 1e-05,
+      "loss": 1.2263851165771484,
+      "step": 8350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26148896469861554,
+      "learning_rate": 1e-05,
+      "loss": 1.2246753692626953,
+      "step": 8360
+    },
+    {
+      "epoch": 0.8810526315789474,
+      "grad_norm": 0.29530702677153525,
+      "learning_rate": 1e-05,
+      "loss": 1.2192961692810058,
+      "step": 8370
+    },
+    {
+      "epoch": 0.8821052631578947,
+      "grad_norm": 0.33107220463825743,
+      "learning_rate": 1e-05,
+      "loss": 1.2515433311462403,
+      "step": 8380
+    },
+    {
+      "epoch": 0.8831578947368421,
+      "grad_norm": 0.2813531777082513,
+      "learning_rate": 1e-05,
+      "loss": 1.2120042800903321,
+      "step": 8390
+    },
+    {
+      "epoch": 0.8842105263157894,
+      "grad_norm": 0.2631012072174228,
+      "learning_rate": 1e-05,
+      "loss": 1.1914825439453125,
+      "step": 8400
+    },
+    {
+      "epoch": 0.8852631578947369,
+      "grad_norm": 0.28064844639259234,
+      "learning_rate": 1e-05,
+      "loss": 1.1918046951293946,
+      "step": 8410
+    },
+    {
+      "epoch": 0.8863157894736842,
+      "grad_norm": 0.2629065991951382,
+      "learning_rate": 1e-05,
+      "loss": 1.1666561126708985,
+      "step": 8420
+    },
+    {
+      "epoch": 0.8873684210526316,
+      "grad_norm": 0.25185093347324305,
+      "learning_rate": 1e-05,
+      "loss": 1.180765151977539,
+      "step": 8430
+    },
+    {
+      "epoch": 0.888421052631579,
+      "grad_norm": 0.24189689406131695,
+      "learning_rate": 1e-05,
+      "loss": 1.1282012939453125,
+      "step": 8440
+    },
+    {
+      "epoch": 0.8894736842105263,
+      "grad_norm": 0.2647004445191544,
+      "learning_rate": 1e-05,
+      "loss": 1.1022151947021483,
+      "step": 8450
+    },
+    {
+      "epoch": 0.8905263157894737,
+      "grad_norm": 0.2641287860444434,
+      "learning_rate": 1e-05,
+      "loss": 1.1134539604187013,
+      "step": 8460
+    },
+    {
+      "epoch": 0.891578947368421,
+      "grad_norm": 0.24069627311122557,
+      "learning_rate": 1e-05,
+      "loss": 1.091534423828125,
+      "step": 8470
+    },
+    {
+      "epoch": 0.8926315789473684,
+      "grad_norm": 0.25115057960806103,
+      "learning_rate": 1e-05,
+      "loss": 1.1029605865478516,
+      "step": 8480
+    },
+    {
+      "epoch": 0.8936842105263157,
+      "grad_norm": 0.2939365023559868,
+      "learning_rate": 1e-05,
+      "loss": 1.0801345825195312,
+      "step": 8490
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.24366124450721913,
+      "learning_rate": 1e-05,
+      "loss": 1.0616696357727051,
+      "step": 8500
+    },
+    {
+      "epoch": 0.8957894736842106,
+      "grad_norm": 0.24286392691962091,
+      "learning_rate": 1e-05,
+      "loss": 1.033091926574707,
+      "step": 8510
+    },
+    {
+      "epoch": 0.8968421052631579,
+      "grad_norm": 0.24644385070077768,
+      "learning_rate": 1e-05,
+      "loss": 1.030784797668457,
+      "step": 8520
+    },
+    {
+      "epoch": 0.8978947368421053,
+      "grad_norm": 0.2560783385054672,
+      "learning_rate": 1e-05,
+      "loss": 1.0137654304504395,
+      "step": 8530
+    },
+    {
+      "epoch": 0.8989473684210526,
+      "grad_norm": 0.2391108699604642,
+      "learning_rate": 1e-05,
+      "loss": 1.006385040283203,
+      "step": 8540
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2377961771517584,
+      "learning_rate": 1e-05,
+      "loss": 0.9708780288696289,
+      "step": 8550
+    },
+    {
+      "epoch": 0.9010526315789473,
+      "grad_norm": 0.2569933743739859,
+      "learning_rate": 1e-05,
+      "loss": 0.9508537292480469,
+      "step": 8560
+    },
+    {
+      "epoch": 0.9021052631578947,
+      "grad_norm": 0.23103241978888703,
+      "learning_rate": 1e-05,
+      "loss": 0.9720380783081055,
+      "step": 8570
+    },
+    {
+      "epoch": 0.9031578947368422,
+      "grad_norm": 0.31694452973073833,
+      "learning_rate": 1e-05,
+      "loss": 0.9239744186401367,
+      "step": 8580
+    },
+    {
+      "epoch": 0.9042105263157895,
+      "grad_norm": 0.2233864432107501,
+      "learning_rate": 1e-05,
+      "loss": 0.9380537986755371,
+      "step": 8590
+    },
+    {
+      "epoch": 0.9052631578947369,
+      "grad_norm": 0.23081261337521336,
+      "learning_rate": 1e-05,
+      "loss": 0.915916633605957,
+      "step": 8600
+    },
+    {
+      "epoch": 0.9063157894736842,
+      "grad_norm": 0.24025606769377367,
+      "learning_rate": 1e-05,
+      "loss": 0.8951493263244629,
+      "step": 8610
+    },
+    {
+      "epoch": 0.9073684210526316,
+      "grad_norm": 0.296841793815868,
+      "learning_rate": 1e-05,
+      "loss": 0.9111743927001953,
+      "step": 8620
+    },
+    {
+      "epoch": 0.9084210526315789,
+      "grad_norm": 0.23098074629664309,
+      "learning_rate": 1e-05,
+      "loss": 0.9580887794494629,
+      "step": 8630
+    },
+    {
+      "epoch": 0.9094736842105263,
+      "grad_norm": 0.22502208082404163,
+      "learning_rate": 1e-05,
+      "loss": 0.8954828262329102,
+      "step": 8640
+    },
+    {
+      "epoch": 0.9105263157894737,
+      "grad_norm": 0.23535685712792298,
+      "learning_rate": 1e-05,
+      "loss": 0.9037555694580078,
+      "step": 8650
+    },
+    {
+      "epoch": 0.911578947368421,
+      "grad_norm": 0.2127950512633359,
+      "learning_rate": 1e-05,
+      "loss": 0.8547103881835938,
+      "step": 8660
+    },
+    {
+      "epoch": 0.9126315789473685,
+      "grad_norm": 0.21882815479778897,
+      "learning_rate": 1e-05,
+      "loss": 0.8527597427368164,
+      "step": 8670
+    },
+    {
+      "epoch": 0.9136842105263158,
+      "grad_norm": 0.23366284804725043,
+      "learning_rate": 1e-05,
+      "loss": 0.876220703125,
+      "step": 8680
+    },
+    {
+      "epoch": 0.9147368421052632,
+      "grad_norm": 0.26444411042998667,
+      "learning_rate": 1e-05,
+      "loss": 0.856623363494873,
+      "step": 8690
+    },
+    {
+      "epoch": 0.9157894736842105,
+      "grad_norm": 0.21715668701179666,
+      "learning_rate": 1e-05,
+      "loss": 0.8695198059082031,
+      "step": 8700
+    },
+    {
+      "epoch": 0.9168421052631579,
+      "grad_norm": 0.22294152638286216,
+      "learning_rate": 1e-05,
+      "loss": 0.8599737167358399,
+      "step": 8710
+    },
+    {
+      "epoch": 0.9178947368421052,
+      "grad_norm": 0.2337711685872282,
+      "learning_rate": 1e-05,
+      "loss": 0.8513544082641602,
+      "step": 8720
+    },
+    {
+      "epoch": 0.9189473684210526,
+      "grad_norm": 0.23906024853957974,
+      "learning_rate": 1e-05,
+      "loss": 0.8523876190185546,
+      "step": 8730
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.22073815599729743,
+      "learning_rate": 1e-05,
+      "loss": 0.8367352485656738,
+      "step": 8740
+    },
+    {
+      "epoch": 0.9210526315789473,
+      "grad_norm": 0.22448568463154753,
+      "learning_rate": 1e-05,
+      "loss": 0.8680167198181152,
+      "step": 8750
+    },
+    {
+      "epoch": 0.9221052631578948,
+      "grad_norm": 0.22246641822129057,
+      "learning_rate": 1e-05,
+      "loss": 0.8439714431762695,
+      "step": 8760
+    },
+    {
+      "epoch": 0.9231578947368421,
+      "grad_norm": 0.23289141837367208,
+      "learning_rate": 1e-05,
+      "loss": 0.8463233947753906,
+      "step": 8770
+    },
+    {
+      "epoch": 0.9242105263157895,
+      "grad_norm": 0.3375630145622655,
+      "learning_rate": 1e-05,
+      "loss": 0.8574393272399903,
+      "step": 8780
+    },
+    {
+      "epoch": 0.9252631578947368,
+      "grad_norm": 0.19651635640156878,
+      "learning_rate": 1e-05,
+      "loss": 0.8035933494567871,
+      "step": 8790
+    },
+    {
+      "epoch": 0.9263157894736842,
+      "grad_norm": 0.24700824062029791,
+      "learning_rate": 1e-05,
+      "loss": 0.7834304809570313,
+      "step": 8800
+    },
+    {
+      "epoch": 0.9273684210526316,
+      "grad_norm": 0.21411575599547564,
+      "learning_rate": 1e-05,
+      "loss": 0.778415298461914,
+      "step": 8810
+    },
+    {
+      "epoch": 0.9284210526315789,
+      "grad_norm": 0.3049708354529429,
+      "learning_rate": 1e-05,
+      "loss": 0.767140245437622,
+      "step": 8820
+    },
+    {
+      "epoch": 0.9294736842105263,
+      "grad_norm": 0.1890049656286049,
+      "learning_rate": 1e-05,
+      "loss": 0.7538487434387207,
+      "step": 8830
+    },
+    {
+      "epoch": 0.9305263157894736,
+      "grad_norm": 0.20811670669103846,
+      "learning_rate": 1e-05,
+      "loss": 0.709379005432129,
+      "step": 8840
+    },
+    {
+      "epoch": 0.9315789473684211,
+      "grad_norm": 0.21457209857479662,
+      "learning_rate": 1e-05,
+      "loss": 0.7053157806396484,
+      "step": 8850
+    },
+    {
+      "epoch": 0.9326315789473684,
+      "grad_norm": 0.19459829081867222,
+      "learning_rate": 1e-05,
+      "loss": 0.698844575881958,
+      "step": 8860
+    },
+    {
+      "epoch": 0.9336842105263158,
+      "grad_norm": 0.21179207207173176,
+      "learning_rate": 1e-05,
+      "loss": 0.687296199798584,
+      "step": 8870
+    },
+    {
+      "epoch": 0.9347368421052632,
+      "grad_norm": 0.1989933845165214,
+      "learning_rate": 1e-05,
+      "loss": 0.6703040122985839,
+      "step": 8880
+    },
+    {
+      "epoch": 0.9357894736842105,
+      "grad_norm": 0.20776682511621972,
+      "learning_rate": 1e-05,
+      "loss": 0.6744589328765869,
+      "step": 8890
+    },
+    {
+      "epoch": 0.9368421052631579,
+      "grad_norm": 0.2092872130461126,
+      "learning_rate": 1e-05,
+      "loss": 0.6761126518249512,
+      "step": 8900
+    },
+    {
+      "epoch": 0.9378947368421052,
+      "grad_norm": 0.22785656675358273,
+      "learning_rate": 1e-05,
+      "loss": 0.675350570678711,
+      "step": 8910
+    },
+    {
+      "epoch": 0.9389473684210526,
+      "grad_norm": 0.21712992946784904,
+      "learning_rate": 1e-05,
+      "loss": 0.6713425636291503,
+      "step": 8920
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23166249573526182,
+      "learning_rate": 1e-05,
+      "loss": 0.6628326416015625,
+      "step": 8930
+    },
+    {
+      "epoch": 0.9410526315789474,
+      "grad_norm": 0.6594057760648365,
+      "learning_rate": 1e-05,
+      "loss": 0.643064546585083,
+      "step": 8940
+    },
+    {
+      "epoch": 0.9421052631578948,
+      "grad_norm": 0.2596828266606076,
+      "learning_rate": 1e-05,
+      "loss": 0.6598286628723145,
+      "step": 8950
+    },
+    {
+      "epoch": 0.9431578947368421,
+      "grad_norm": 0.19724597614625558,
+      "learning_rate": 1e-05,
+      "loss": 0.5980873107910156,
+      "step": 8960
+    },
+    {
+      "epoch": 0.9442105263157895,
+      "grad_norm": 0.2078811918029264,
+      "learning_rate": 1e-05,
+      "loss": 0.6211897850036621,
+      "step": 8970
+    },
+    {
+      "epoch": 0.9452631578947368,
+      "grad_norm": 0.21188684099627883,
+      "learning_rate": 1e-05,
+      "loss": 0.5906908988952637,
+      "step": 8980
+    },
+    {
+      "epoch": 0.9463157894736842,
+      "grad_norm": 0.20028773107785666,
+      "learning_rate": 1e-05,
+      "loss": 0.5580528259277344,
+      "step": 8990
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.18223006968990166,
+      "learning_rate": 1e-05,
+      "loss": 0.5851156711578369,
+      "step": 9000
+    },
+    {
+      "epoch": 0.9484210526315789,
+      "grad_norm": 0.19661293931890084,
+      "learning_rate": 1e-05,
+      "loss": 0.5192183494567871,
+      "step": 9010
+    },
+    {
+      "epoch": 0.9494736842105264,
+      "grad_norm": 0.18076896046499094,
+      "learning_rate": 1e-05,
+      "loss": 0.513965654373169,
+      "step": 9020
+    },
+    {
+      "epoch": 0.9505263157894737,
+      "grad_norm": 0.18695006786265717,
+      "learning_rate": 1e-05,
+      "loss": 0.5101747989654541,
+      "step": 9030
+    },
+    {
+      "epoch": 0.9515789473684211,
+      "grad_norm": 0.20248565390516432,
+      "learning_rate": 1e-05,
+      "loss": 0.5146926879882813,
+      "step": 9040
+    },
+    {
+      "epoch": 0.9526315789473684,
+      "grad_norm": 0.17387515897766911,
+      "learning_rate": 1e-05,
+      "loss": 0.5397676467895508,
+      "step": 9050
+    },
+    {
+      "epoch": 0.9536842105263158,
+      "grad_norm": 0.17774197277066184,
+      "learning_rate": 1e-05,
+      "loss": 0.5814432144165039,
+      "step": 9060
+    },
+    {
+      "epoch": 0.9547368421052631,
+      "grad_norm": 0.19735269435711078,
+      "learning_rate": 1e-05,
+      "loss": 0.5320480346679688,
+      "step": 9070
+    },
+    {
+      "epoch": 0.9557894736842105,
+      "grad_norm": 0.26012497009104923,
+      "learning_rate": 1e-05,
+      "loss": 0.48093390464782715,
+      "step": 9080
+    },
+    {
+      "epoch": 0.9568421052631579,
+      "grad_norm": 0.18532656117120772,
+      "learning_rate": 1e-05,
+      "loss": 0.5239907741546631,
+      "step": 9090
+    },
+    {
+      "epoch": 0.9578947368421052,
+      "grad_norm": 0.20636730673147705,
+      "learning_rate": 1e-05,
+      "loss": 0.5028749942779541,
+      "step": 9100
+    },
+    {
+      "epoch": 0.9589473684210527,
+      "grad_norm": 0.19865376703305168,
+      "learning_rate": 1e-05,
+      "loss": 0.47848663330078123,
+      "step": 9110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.17207509308168797,
+      "learning_rate": 1e-05,
+      "loss": 0.46389479637145997,
+      "step": 9120
+    },
+    {
+      "epoch": 0.9610526315789474,
+      "grad_norm": 0.20150743395902215,
+      "learning_rate": 1e-05,
+      "loss": 0.45012102127075193,
+      "step": 9130
+    },
+    {
+      "epoch": 0.9621052631578947,
+      "grad_norm": 0.30859210069285076,
+      "learning_rate": 1e-05,
+      "loss": 0.4895476818084717,
+      "step": 9140
+    },
+    {
+      "epoch": 0.9631578947368421,
+      "grad_norm": 0.20057802579831513,
+      "learning_rate": 1e-05,
+      "loss": 0.45813713073730467,
+      "step": 9150
+    },
+    {
+      "epoch": 0.9642105263157895,
+      "grad_norm": 0.2355090125924601,
+      "learning_rate": 1e-05,
+      "loss": 0.4826399803161621,
+      "step": 9160
+    },
+    {
+      "epoch": 0.9652631578947368,
+      "grad_norm": 0.17696274625002115,
+      "learning_rate": 1e-05,
+      "loss": 0.46065683364868165,
+      "step": 9170
+    },
+    {
+      "epoch": 0.9663157894736842,
+      "grad_norm": 0.18036750784869485,
+      "learning_rate": 1e-05,
+      "loss": 0.4703228950500488,
+      "step": 9180
+    },
+    {
+      "epoch": 0.9673684210526315,
+      "grad_norm": 0.19082468335595296,
+      "learning_rate": 1e-05,
+      "loss": 0.4499660491943359,
+      "step": 9190
+    },
+    {
+      "epoch": 0.968421052631579,
+      "grad_norm": 0.2556638794769354,
+      "learning_rate": 1e-05,
+      "loss": 0.4480300426483154,
+      "step": 9200
+    },
+    {
+      "epoch": 0.9694736842105263,
+      "grad_norm": 0.17304593928985593,
+      "learning_rate": 1e-05,
+      "loss": 0.44375810623168943,
+      "step": 9210
+    },
+    {
+      "epoch": 0.9705263157894737,
+      "grad_norm": 0.21181791929041688,
+      "learning_rate": 1e-05,
+      "loss": 0.41954312324523924,
+      "step": 9220
+    },
+    {
+      "epoch": 0.9715789473684211,
+      "grad_norm": 0.17140833116108528,
+      "learning_rate": 1e-05,
+      "loss": 0.48293113708496094,
+      "step": 9230
+    },
+    {
+      "epoch": 0.9726315789473684,
+      "grad_norm": 0.168770059689768,
+      "learning_rate": 1e-05,
+      "loss": 0.46772146224975586,
+      "step": 9240
+    },
+    {
+      "epoch": 0.9736842105263158,
+      "grad_norm": 0.1796396146615267,
+      "learning_rate": 1e-05,
+      "loss": 0.5024192333221436,
+      "step": 9250
+    },
+    {
+      "epoch": 0.9747368421052631,
+      "grad_norm": 0.2149394400309839,
+      "learning_rate": 1e-05,
+      "loss": 0.5203466415405273,
+      "step": 9260
+    },
+    {
+      "epoch": 0.9757894736842105,
+      "grad_norm": 0.19593807462401958,
+      "learning_rate": 1e-05,
+      "loss": 0.5324414253234864,
+      "step": 9270
+    },
+    {
+      "epoch": 0.9768421052631578,
+      "grad_norm": 0.17433712308479005,
+      "learning_rate": 1e-05,
+      "loss": 0.513987684249878,
+      "step": 9280
+    },
+    {
+      "epoch": 0.9778947368421053,
+      "grad_norm": 0.37517583171073404,
+      "learning_rate": 1e-05,
+      "loss": 0.5582686424255371,
+      "step": 9290
+    },
+    {
+      "epoch": 0.9789473684210527,
+      "grad_norm": 0.20079454881211847,
+      "learning_rate": 1e-05,
+      "loss": 0.5541967868804931,
+      "step": 9300
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1918408639036845,
+      "learning_rate": 1e-05,
+      "loss": 0.5959220886230469,
+      "step": 9310
+    },
+    {
+      "epoch": 0.9810526315789474,
+      "grad_norm": 0.1886389073090896,
+      "learning_rate": 1e-05,
+      "loss": 0.6560276031494141,
+      "step": 9320
+    },
+    {
+      "epoch": 0.9821052631578947,
+      "grad_norm": 0.1978670830203482,
+      "learning_rate": 1e-05,
+      "loss": 0.6445225238800049,
+      "step": 9330
+    },
+    {
+      "epoch": 0.9831578947368421,
+      "grad_norm": 0.5227765274459174,
+      "learning_rate": 1e-05,
+      "loss": 0.6521971702575684,
+      "step": 9340
+    },
+    {
+      "epoch": 0.9842105263157894,
+      "grad_norm": 0.2339773380825472,
+      "learning_rate": 1e-05,
+      "loss": 0.6857166290283203,
+      "step": 9350
+    },
+    {
+      "epoch": 0.9852631578947368,
+      "grad_norm": 0.2281839158566163,
+      "learning_rate": 1e-05,
+      "loss": 0.7210614204406738,
+      "step": 9360
+    },
+    {
+      "epoch": 0.9863157894736843,
+      "grad_norm": 0.22494691818750182,
+      "learning_rate": 1e-05,
+      "loss": 0.7327334880828857,
+      "step": 9370
+    },
+    {
+      "epoch": 0.9873684210526316,
+      "grad_norm": 0.20170374103708558,
+      "learning_rate": 1e-05,
+      "loss": 0.7342099189758301,
+      "step": 9380
+    },
+    {
+      "epoch": 0.988421052631579,
+      "grad_norm": 0.21994834523949286,
+      "learning_rate": 1e-05,
+      "loss": 0.7164675712585449,
+      "step": 9390
+    },
+    {
+      "epoch": 0.9894736842105263,
+      "grad_norm": 0.21077513656316527,
+      "learning_rate": 1e-05,
+      "loss": 0.774052619934082,
+      "step": 9400
+    },
+    {
+      "epoch": 0.9905263157894737,
+      "grad_norm": 0.21697600028260566,
+      "learning_rate": 1e-05,
+      "loss": 0.7530053138732911,
+      "step": 9410
+    },
+    {
+      "epoch": 0.991578947368421,
+      "grad_norm": 0.22937584136651137,
+      "learning_rate": 1e-05,
+      "loss": 0.777883243560791,
+      "step": 9420
+    },
+    {
+      "epoch": 0.9926315789473684,
+      "grad_norm": 0.21062351703765386,
+      "learning_rate": 1e-05,
+      "loss": 0.7843510627746582,
+      "step": 9430
+    },
+    {
+      "epoch": 0.9936842105263158,
+      "grad_norm": 0.20806165835478788,
+      "learning_rate": 1e-05,
+      "loss": 0.7711593627929687,
+      "step": 9440
+    },
+    {
+      "epoch": 0.9947368421052631,
+      "grad_norm": 0.20677131681224006,
+      "learning_rate": 1e-05,
+      "loss": 0.7815569400787353,
+      "step": 9450
+    },
+    {
+      "epoch": 0.9957894736842106,
+      "grad_norm": 0.20123911148814108,
+      "learning_rate": 1e-05,
+      "loss": 0.7943475246429443,
+      "step": 9460
+    },
+    {
+      "epoch": 0.9968421052631579,
+      "grad_norm": 0.22764163198173562,
+      "learning_rate": 1e-05,
+      "loss": 0.8312942504882812,
+      "step": 9470
+    },
+    {
+      "epoch": 0.9978947368421053,
+      "grad_norm": 0.3033912254170302,
+      "learning_rate": 1e-05,
+      "loss": 0.8159421920776367,
+      "step": 9480
+    },
+    {
+      "epoch": 0.9989473684210526,
+      "grad_norm": 0.21777961853289218,
+      "learning_rate": 1e-05,
+      "loss": 0.8064357757568359,
+      "step": 9490
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.22407410341960468,
+      "learning_rate": 1e-05,
+      "loss": 0.7963273048400878,
+      "step": 9500
+    },
+    {
+      "epoch": 1.0,
+      "step": 9500,
+      "total_flos": 1.7566495445876736e+17,
+      "train_loss": 0.4299021988918907,
+      "train_runtime": 81916.7452,
+      "train_samples_per_second": 14.844,
+      "train_steps_per_second": 0.116
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 9500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.7566495445876736e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}