diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,18458 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.800060794893229,
+  "eval_steps": 500,
+  "global_step": 2632,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00030397446614484385,
+      "grad_norm": 23.652748107910156,
+      "learning_rate": 0.0,
+      "loss": 6.191377639770508,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006079489322896877,
+      "grad_norm": 19.74618911743164,
+      "learning_rate": 2.9999999999999997e-06,
+      "loss": 6.255431175231934,
+      "step": 2
+    },
+    {
+      "epoch": 0.0009119233984345315,
+      "grad_norm": 16.828752517700195,
+      "learning_rate": 5.999999999999999e-06,
+      "loss": 6.320460796356201,
+      "step": 3
+    },
+    {
+      "epoch": 0.0012158978645793754,
+      "grad_norm": 13.639242172241211,
+      "learning_rate": 8.999999999999999e-06,
+      "loss": 6.266885757446289,
+      "step": 4
+    },
+    {
+      "epoch": 0.001519872330724219,
+      "grad_norm": 10.231013298034668,
+      "learning_rate": 1.1999999999999999e-05,
+      "loss": 6.487092018127441,
+      "step": 5
+    },
+    {
+      "epoch": 0.001823846796869063,
+      "grad_norm": 13.016983985900879,
+      "learning_rate": 1.4999999999999999e-05,
+      "loss": 5.850100994110107,
+      "step": 6
+    },
+    {
+      "epoch": 0.002127821263013907,
+      "grad_norm": 12.495027542114258,
+      "learning_rate": 1.7999999999999997e-05,
+      "loss": 6.317249298095703,
+      "step": 7
+    },
+    {
+      "epoch": 0.002431795729158751,
+      "grad_norm": 9.531314849853516,
+      "learning_rate": 2.1e-05,
+      "loss": 6.295767784118652,
+      "step": 8
+    },
+    {
+      "epoch": 0.0027357701953035947,
+      "grad_norm": 10.633639335632324,
+      "learning_rate": 2.3999999999999997e-05,
+      "loss": 5.924046516418457,
+      "step": 9
+    },
+    {
+      "epoch": 0.003039744661448438,
+      "grad_norm": 9.606488227844238,
+      "learning_rate": 2.6999999999999996e-05,
+      "loss": 5.981108665466309,
+      "step": 10
+    },
+    {
+      "epoch": 0.003343719127593282,
+      "grad_norm": 16.802595138549805,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 5.815858364105225,
+      "step": 11
+    },
+    {
+      "epoch": 0.003647693593738126,
+      "grad_norm": 9.00236701965332,
+      "learning_rate": 3.2999999999999996e-05,
+      "loss": 5.804163932800293,
+      "step": 12
+    },
+    {
+      "epoch": 0.0039516680598829694,
+      "grad_norm": 6.464416027069092,
+      "learning_rate": 3.5999999999999994e-05,
+      "loss": 5.80509614944458,
+      "step": 13
+    },
+    {
+      "epoch": 0.004255642526027814,
+      "grad_norm": 7.549493312835693,
+      "learning_rate": 3.9e-05,
+      "loss": 5.750925064086914,
+      "step": 14
+    },
+    {
+      "epoch": 0.004559616992172657,
+      "grad_norm": 15.18075180053711,
+      "learning_rate": 4.2e-05,
+      "loss": 6.196191787719727,
+      "step": 15
+    },
+    {
+      "epoch": 0.004863591458317502,
+      "grad_norm": 6.047942161560059,
+      "learning_rate": 4.4999999999999996e-05,
+      "loss": 5.7587480545043945,
+      "step": 16
+    },
+    {
+      "epoch": 0.005167565924462345,
+      "grad_norm": 6.147083759307861,
+      "learning_rate": 4.7999999999999994e-05,
+      "loss": 5.732815742492676,
+      "step": 17
+    },
+    {
+      "epoch": 0.005471540390607189,
+      "grad_norm": 9.685389518737793,
+      "learning_rate": 5.1e-05,
+      "loss": 6.04966926574707,
+      "step": 18
+    },
+    {
+      "epoch": 0.005775514856752033,
+      "grad_norm": 5.859086513519287,
+      "learning_rate": 5.399999999999999e-05,
+      "loss": 5.353918552398682,
+      "step": 19
+    },
+    {
+      "epoch": 0.006079489322896876,
+      "grad_norm": 18.99468994140625,
+      "learning_rate": 5.6999999999999996e-05,
+      "loss": 5.736433029174805,
+      "step": 20
+    },
+    {
+      "epoch": 0.006383463789041721,
+      "grad_norm": 5.2469563484191895,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 5.593536853790283,
+      "step": 21
+    },
+    {
+      "epoch": 0.006687438255186564,
+      "grad_norm": 5.316288948059082,
+      "learning_rate": 6.299999999999999e-05,
+      "loss": 5.878802299499512,
+      "step": 22
+    },
+    {
+      "epoch": 0.0069914127213314085,
+      "grad_norm": 5.482022285461426,
+      "learning_rate": 6.599999999999999e-05,
+      "loss": 5.765041828155518,
+      "step": 23
+    },
+    {
+      "epoch": 0.007295387187476252,
+      "grad_norm": 5.783045291900635,
+      "learning_rate": 6.9e-05,
+      "loss": 5.709415912628174,
+      "step": 24
+    },
+    {
+      "epoch": 0.007599361653621095,
+      "grad_norm": 5.530306816101074,
+      "learning_rate": 7.199999999999999e-05,
+      "loss": 5.4808349609375,
+      "step": 25
+    },
+    {
+      "epoch": 0.007903336119765939,
+      "grad_norm": 5.633169651031494,
+      "learning_rate": 7.5e-05,
+      "loss": 5.796350002288818,
+      "step": 26
+    },
+    {
+      "epoch": 0.008207310585910784,
+      "grad_norm": 5.07309627532959,
+      "learning_rate": 7.8e-05,
+      "loss": 5.728445053100586,
+      "step": 27
+    },
+    {
+      "epoch": 0.008511285052055628,
+      "grad_norm": 4.835383415222168,
+      "learning_rate": 8.1e-05,
+      "loss": 5.440605163574219,
+      "step": 28
+    },
+    {
+      "epoch": 0.008815259518200471,
+      "grad_norm": 4.487118721008301,
+      "learning_rate": 8.4e-05,
+      "loss": 5.701571941375732,
+      "step": 29
+    },
+    {
+      "epoch": 0.009119233984345314,
+      "grad_norm": 5.415592670440674,
+      "learning_rate": 8.699999999999999e-05,
+      "loss": 5.756715774536133,
+      "step": 30
+    },
+    {
+      "epoch": 0.00942320845049016,
+      "grad_norm": 3.7513833045959473,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 5.869050979614258,
+      "step": 31
+    },
+    {
+      "epoch": 0.009727182916635003,
+      "grad_norm": 7.317835330963135,
+      "learning_rate": 9.3e-05,
+      "loss": 5.736352920532227,
+      "step": 32
+    },
+    {
+      "epoch": 0.010031157382779847,
+      "grad_norm": 7.708019733428955,
+      "learning_rate": 9.599999999999999e-05,
+      "loss": 5.661643981933594,
+      "step": 33
+    },
+    {
+      "epoch": 0.01033513184892469,
+      "grad_norm": 4.413952350616455,
+      "learning_rate": 9.9e-05,
+      "loss": 5.5280866622924805,
+      "step": 34
+    },
+    {
+      "epoch": 0.010639106315069534,
+      "grad_norm": 4.912237167358398,
+      "learning_rate": 0.000102,
+      "loss": 5.864815711975098,
+      "step": 35
+    },
+    {
+      "epoch": 0.010943080781214379,
+      "grad_norm": 7.133541584014893,
+      "learning_rate": 0.00010499999999999999,
+      "loss": 5.251829147338867,
+      "step": 36
+    },
+    {
+      "epoch": 0.011247055247359222,
+      "grad_norm": 5.643782615661621,
+      "learning_rate": 0.00010799999999999998,
+      "loss": 5.728849411010742,
+      "step": 37
+    },
+    {
+      "epoch": 0.011551029713504066,
+      "grad_norm": 4.055923938751221,
+      "learning_rate": 0.00011099999999999999,
+      "loss": 5.580687999725342,
+      "step": 38
+    },
+    {
+      "epoch": 0.01185500417964891,
+      "grad_norm": 4.0418291091918945,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 5.548113822937012,
+      "step": 39
+    },
+    {
+      "epoch": 0.012158978645793753,
+      "grad_norm": 4.151160717010498,
+      "learning_rate": 0.000117,
+      "loss": 5.519264221191406,
+      "step": 40
+    },
+    {
+      "epoch": 0.012462953111938598,
+      "grad_norm": 5.462052345275879,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 5.478485107421875,
+      "step": 41
+    },
+    {
+      "epoch": 0.012766927578083441,
+      "grad_norm": 4.804545879364014,
+      "learning_rate": 0.00012299999999999998,
+      "loss": 5.6366472244262695,
+      "step": 42
+    },
+    {
+      "epoch": 0.013070902044228285,
+      "grad_norm": 3.977419376373291,
+      "learning_rate": 0.00012599999999999997,
+      "loss": 5.5130815505981445,
+      "step": 43
+    },
+    {
+      "epoch": 0.013374876510373128,
+      "grad_norm": 4.459465026855469,
+      "learning_rate": 0.000129,
+      "loss": 5.677802085876465,
+      "step": 44
+    },
+    {
+      "epoch": 0.013678850976517972,
+      "grad_norm": 4.534205436706543,
+      "learning_rate": 0.00013199999999999998,
+      "loss": 5.673933506011963,
+      "step": 45
+    },
+    {
+      "epoch": 0.013982825442662817,
+      "grad_norm": 3.907681465148926,
+      "learning_rate": 0.000135,
+      "loss": 5.693915367126465,
+      "step": 46
+    },
+    {
+      "epoch": 0.01428679990880766,
+      "grad_norm": 4.340635776519775,
+      "learning_rate": 0.000138,
+      "loss": 5.715434551239014,
+      "step": 47
+    },
+    {
+      "epoch": 0.014590774374952504,
+      "grad_norm": 4.356265544891357,
+      "learning_rate": 0.00014099999999999998,
+      "loss": 5.990167140960693,
+      "step": 48
+    },
+    {
+      "epoch": 0.014894748841097347,
+      "grad_norm": 5.513248443603516,
+      "learning_rate": 0.00014399999999999998,
+      "loss": 5.891347408294678,
+      "step": 49
+    },
+    {
+      "epoch": 0.01519872330724219,
+      "grad_norm": 4.634105205535889,
+      "learning_rate": 0.000147,
+      "loss": 5.806520462036133,
+      "step": 50
+    },
+    {
+      "epoch": 0.015502697773387036,
+      "grad_norm": 3.3471977710723877,
+      "learning_rate": 0.00015,
+      "loss": 5.486179828643799,
+      "step": 51
+    },
+    {
+      "epoch": 0.015806672239531878,
+      "grad_norm": 4.021389484405518,
+      "learning_rate": 0.00015,
+      "loss": 5.58123254776001,
+      "step": 52
+    },
+    {
+      "epoch": 0.016110646705676725,
+      "grad_norm": 4.883877754211426,
+      "learning_rate": 0.00015,
+      "loss": 5.569307327270508,
+      "step": 53
+    },
+    {
+      "epoch": 0.016414621171821568,
+      "grad_norm": 4.175922870635986,
+      "learning_rate": 0.00015,
+      "loss": 5.504737854003906,
+      "step": 54
+    },
+    {
+      "epoch": 0.01671859563796641,
+      "grad_norm": 4.9562530517578125,
+      "learning_rate": 0.00015,
+      "loss": 5.4153032302856445,
+      "step": 55
+    },
+    {
+      "epoch": 0.017022570104111255,
+      "grad_norm": 5.975865364074707,
+      "learning_rate": 0.00015,
+      "loss": 5.386896133422852,
+      "step": 56
+    },
+    {
+      "epoch": 0.0173265445702561,
+      "grad_norm": 2.985295057296753,
+      "learning_rate": 0.00015,
+      "loss": 5.608443260192871,
+      "step": 57
+    },
+    {
+      "epoch": 0.017630519036400942,
+      "grad_norm": 4.432838439941406,
+      "learning_rate": 0.00015,
+      "loss": 5.7831878662109375,
+      "step": 58
+    },
+    {
+      "epoch": 0.017934493502545786,
+      "grad_norm": 5.696512699127197,
+      "learning_rate": 0.00015,
+      "loss": 5.533647537231445,
+      "step": 59
+    },
+    {
+      "epoch": 0.01823846796869063,
+      "grad_norm": 3.7115120887756348,
+      "learning_rate": 0.00015,
+      "loss": 5.229396343231201,
+      "step": 60
+    },
+    {
+      "epoch": 0.018542442434835472,
+      "grad_norm": 3.559370517730713,
+      "learning_rate": 0.00015,
+      "loss": 5.321964263916016,
+      "step": 61
+    },
+    {
+      "epoch": 0.01884641690098032,
+      "grad_norm": 4.85106897354126,
+      "learning_rate": 0.00015,
+      "loss": 5.475746154785156,
+      "step": 62
+    },
+    {
+      "epoch": 0.019150391367125163,
+      "grad_norm": 4.2979817390441895,
+      "learning_rate": 0.00015,
+      "loss": 5.459148406982422,
+      "step": 63
+    },
+    {
+      "epoch": 0.019454365833270006,
+      "grad_norm": 3.3267762660980225,
+      "learning_rate": 0.00015,
+      "loss": 5.5134735107421875,
+      "step": 64
+    },
+    {
+      "epoch": 0.01975834029941485,
+      "grad_norm": 3.584444046020508,
+      "learning_rate": 0.00015,
+      "loss": 5.310276985168457,
+      "step": 65
+    },
+    {
+      "epoch": 0.020062314765559693,
+      "grad_norm": 3.6489877700805664,
+      "learning_rate": 0.00015,
+      "loss": 5.636244773864746,
+      "step": 66
+    },
+    {
+      "epoch": 0.020366289231704537,
+      "grad_norm": 4.361774921417236,
+      "learning_rate": 0.00015,
+      "loss": 5.4975433349609375,
+      "step": 67
+    },
+    {
+      "epoch": 0.02067026369784938,
+      "grad_norm": 4.080193519592285,
+      "learning_rate": 0.00015,
+      "loss": 5.8029255867004395,
+      "step": 68
+    },
+    {
+      "epoch": 0.020974238163994224,
+      "grad_norm": 4.703793048858643,
+      "learning_rate": 0.00015,
+      "loss": 5.6140289306640625,
+      "step": 69
+    },
+    {
+      "epoch": 0.021278212630139067,
+      "grad_norm": 3.6464405059814453,
+      "learning_rate": 0.00015,
+      "loss": 5.440647125244141,
+      "step": 70
+    },
+    {
+      "epoch": 0.02158218709628391,
+      "grad_norm": 4.654208183288574,
+      "learning_rate": 0.00015,
+      "loss": 5.562455177307129,
+      "step": 71
+    },
+    {
+      "epoch": 0.021886161562428758,
+      "grad_norm": 4.247203350067139,
+      "learning_rate": 0.00015,
+      "loss": 5.615410804748535,
+      "step": 72
+    },
+    {
+      "epoch": 0.0221901360285736,
+      "grad_norm": 3.5184624195098877,
+      "learning_rate": 0.00015,
+      "loss": 5.302039623260498,
+      "step": 73
+    },
+    {
+      "epoch": 0.022494110494718444,
+      "grad_norm": 5.657173156738281,
+      "learning_rate": 0.00015,
+      "loss": 6.035904884338379,
+      "step": 74
+    },
+    {
+      "epoch": 0.022798084960863288,
+      "grad_norm": 4.342555522918701,
+      "learning_rate": 0.00015,
+      "loss": 5.6045732498168945,
+      "step": 75
+    },
+    {
+      "epoch": 0.02310205942700813,
+      "grad_norm": 4.2890753746032715,
+      "learning_rate": 0.00015,
+      "loss": 5.720365047454834,
+      "step": 76
+    },
+    {
+      "epoch": 0.023406033893152975,
+      "grad_norm": 3.5670220851898193,
+      "learning_rate": 0.00015,
+      "loss": 5.492122650146484,
+      "step": 77
+    },
+    {
+      "epoch": 0.02371000835929782,
+      "grad_norm": 2.5429587364196777,
+      "learning_rate": 0.00015,
+      "loss": 5.509317398071289,
+      "step": 78
+    },
+    {
+      "epoch": 0.024013982825442662,
+      "grad_norm": 3.532214879989624,
+      "learning_rate": 0.00015,
+      "loss": 5.711646556854248,
+      "step": 79
+    },
+    {
+      "epoch": 0.024317957291587505,
+      "grad_norm": 3.2889177799224854,
+      "learning_rate": 0.00015,
+      "loss": 5.182368278503418,
+      "step": 80
+    },
+    {
+      "epoch": 0.02462193175773235,
+      "grad_norm": 3.4412119388580322,
+      "learning_rate": 0.00015,
+      "loss": 5.374791145324707,
+      "step": 81
+    },
+    {
+      "epoch": 0.024925906223877196,
+      "grad_norm": 3.066089630126953,
+      "learning_rate": 0.00015,
+      "loss": 5.340858459472656,
+      "step": 82
+    },
+    {
+      "epoch": 0.02522988069002204,
+      "grad_norm": 4.015556812286377,
+      "learning_rate": 0.00015,
+      "loss": 5.2786054611206055,
+      "step": 83
+    },
+    {
+      "epoch": 0.025533855156166883,
+      "grad_norm": 3.423524856567383,
+      "learning_rate": 0.00015,
+      "loss": 5.485323905944824,
+      "step": 84
+    },
+    {
+      "epoch": 0.025837829622311726,
+      "grad_norm": 2.9315497875213623,
+      "learning_rate": 0.00015,
+      "loss": 5.255699157714844,
+      "step": 85
+    },
+    {
+      "epoch": 0.02614180408845657,
+      "grad_norm": 4.309633255004883,
+      "learning_rate": 0.00015,
+      "loss": 5.466099739074707,
+      "step": 86
+    },
+    {
+      "epoch": 0.026445778554601413,
+      "grad_norm": 3.510117530822754,
+      "learning_rate": 0.00015,
+      "loss": 5.021357536315918,
+      "step": 87
+    },
+    {
+      "epoch": 0.026749753020746257,
+      "grad_norm": 3.253308057785034,
+      "learning_rate": 0.00015,
+      "loss": 4.996673107147217,
+      "step": 88
+    },
+    {
+      "epoch": 0.0270537274868911,
+      "grad_norm": 3.117910146713257,
+      "learning_rate": 0.00015,
+      "loss": 5.673067092895508,
+      "step": 89
+    },
+    {
+      "epoch": 0.027357701953035943,
+      "grad_norm": 5.433970928192139,
+      "learning_rate": 0.00015,
+      "loss": 5.477117538452148,
+      "step": 90
+    },
+    {
+      "epoch": 0.02766167641918079,
+      "grad_norm": 3.7955410480499268,
+      "learning_rate": 0.00015,
+      "loss": 5.475001811981201,
+      "step": 91
+    },
+    {
+      "epoch": 0.027965650885325634,
+      "grad_norm": 3.293415069580078,
+      "learning_rate": 0.00015,
+      "loss": 5.464406967163086,
+      "step": 92
+    },
+    {
+      "epoch": 0.028269625351470477,
+      "grad_norm": 2.9583168029785156,
+      "learning_rate": 0.00015,
+      "loss": 5.328250885009766,
+      "step": 93
+    },
+    {
+      "epoch": 0.02857359981761532,
+      "grad_norm": 4.110863208770752,
+      "learning_rate": 0.00015,
+      "loss": 5.267874717712402,
+      "step": 94
+    },
+    {
+      "epoch": 0.028877574283760164,
+      "grad_norm": 3.1186840534210205,
+      "learning_rate": 0.00015,
+      "loss": 5.500239849090576,
+      "step": 95
+    },
+    {
+      "epoch": 0.029181548749905008,
+      "grad_norm": 4.286520957946777,
+      "learning_rate": 0.00015,
+      "loss": 5.513062477111816,
+      "step": 96
+    },
+    {
+      "epoch": 0.02948552321604985,
+      "grad_norm": 3.281066656112671,
+      "learning_rate": 0.00015,
+      "loss": 5.406864643096924,
+      "step": 97
+    },
+    {
+      "epoch": 0.029789497682194695,
+      "grad_norm": 3.563976526260376,
+      "learning_rate": 0.00015,
+      "loss": 5.272912979125977,
+      "step": 98
+    },
+    {
+      "epoch": 0.030093472148339538,
+      "grad_norm": 3.4657142162323,
+      "learning_rate": 0.00015,
+      "loss": 5.352265357971191,
+      "step": 99
+    },
+    {
+      "epoch": 0.03039744661448438,
+      "grad_norm": 3.2147998809814453,
+      "learning_rate": 0.00015,
+      "loss": 5.4894819259643555,
+      "step": 100
+    },
+    {
+      "epoch": 0.03070142108062923,
+      "grad_norm": 3.992326259613037,
+      "learning_rate": 0.00015,
+      "loss": 5.59952449798584,
+      "step": 101
+    },
+    {
+      "epoch": 0.031005395546774072,
+      "grad_norm": 3.4759645462036133,
+      "learning_rate": 0.00015,
+      "loss": 5.554427146911621,
+      "step": 102
+    },
+    {
+      "epoch": 0.03130937001291891,
+      "grad_norm": 3.2231838703155518,
+      "learning_rate": 0.00015,
+      "loss": 5.271030426025391,
+      "step": 103
+    },
+    {
+      "epoch": 0.031613344479063755,
+      "grad_norm": 2.951690435409546,
+      "learning_rate": 0.00015,
+      "loss": 5.718379020690918,
+      "step": 104
+    },
+    {
+      "epoch": 0.031917318945208606,
+      "grad_norm": 3.488664150238037,
+      "learning_rate": 0.00015,
+      "loss": 5.122199058532715,
+      "step": 105
+    },
+    {
+      "epoch": 0.03222129341135345,
+      "grad_norm": 3.0927844047546387,
+      "learning_rate": 0.00015,
+      "loss": 5.1446967124938965,
+      "step": 106
+    },
+    {
+      "epoch": 0.03252526787749829,
+      "grad_norm": 3.090449810028076,
+      "learning_rate": 0.00015,
+      "loss": 5.014104843139648,
+      "step": 107
+    },
+    {
+      "epoch": 0.032829242343643136,
+      "grad_norm": 3.4854650497436523,
+      "learning_rate": 0.00015,
+      "loss": 5.356331825256348,
+      "step": 108
+    },
+    {
+      "epoch": 0.03313321680978798,
+      "grad_norm": 2.3135223388671875,
+      "learning_rate": 0.00015,
+      "loss": 5.280098915100098,
+      "step": 109
+    },
+    {
+      "epoch": 0.03343719127593282,
+      "grad_norm": 2.6215646266937256,
+      "learning_rate": 0.00015,
+      "loss": 5.178532600402832,
+      "step": 110
+    },
+    {
+      "epoch": 0.03374116574207767,
+      "grad_norm": 4.62375545501709,
+      "learning_rate": 0.00015,
+      "loss": 5.255846977233887,
+      "step": 111
+    },
+    {
+      "epoch": 0.03404514020822251,
+      "grad_norm": 2.717853307723999,
+      "learning_rate": 0.00015,
+      "loss": 5.31281852722168,
+      "step": 112
+    },
+    {
+      "epoch": 0.034349114674367354,
+      "grad_norm": 2.631622791290283,
+      "learning_rate": 0.00015,
+      "loss": 5.4918012619018555,
+      "step": 113
+    },
+    {
+      "epoch": 0.0346530891405122,
+      "grad_norm": 4.237459182739258,
+      "learning_rate": 0.00015,
+      "loss": 5.828828811645508,
+      "step": 114
+    },
+    {
+      "epoch": 0.03495706360665704,
+      "grad_norm": 2.689192533493042,
+      "learning_rate": 0.00015,
+      "loss": 5.428500175476074,
+      "step": 115
+    },
+    {
+      "epoch": 0.035261038072801884,
+      "grad_norm": 3.2305634021759033,
+      "learning_rate": 0.00015,
+      "loss": 5.279057025909424,
+      "step": 116
+    },
+    {
+      "epoch": 0.03556501253894673,
+      "grad_norm": 2.967231035232544,
+      "learning_rate": 0.00015,
+      "loss": 5.206664562225342,
+      "step": 117
+    },
+    {
+      "epoch": 0.03586898700509157,
+      "grad_norm": 3.3685953617095947,
+      "learning_rate": 0.00015,
+      "loss": 5.523316383361816,
+      "step": 118
+    },
+    {
+      "epoch": 0.036172961471236414,
+      "grad_norm": 3.004302740097046,
+      "learning_rate": 0.00015,
+      "loss": 5.481061935424805,
+      "step": 119
+    },
+    {
+      "epoch": 0.03647693593738126,
+      "grad_norm": 2.6062612533569336,
+      "learning_rate": 0.00015,
+      "loss": 5.400457382202148,
+      "step": 120
+    },
+    {
+      "epoch": 0.0367809104035261,
+      "grad_norm": 2.475714921951294,
+      "learning_rate": 0.00015,
+      "loss": 5.417738914489746,
+      "step": 121
+    },
+    {
+      "epoch": 0.037084884869670945,
+      "grad_norm": 2.9416444301605225,
+      "learning_rate": 0.00015,
+      "loss": 5.242222785949707,
+      "step": 122
+    },
+    {
+      "epoch": 0.03738885933581579,
+      "grad_norm": 2.6816933155059814,
+      "learning_rate": 0.00015,
+      "loss": 5.048537254333496,
+      "step": 123
+    },
+    {
+      "epoch": 0.03769283380196064,
+      "grad_norm": 3.229828119277954,
+      "learning_rate": 0.00015,
+      "loss": 5.067062854766846,
+      "step": 124
+    },
+    {
+      "epoch": 0.03799680826810548,
+      "grad_norm": 2.75730037689209,
+      "learning_rate": 0.00015,
+      "loss": 5.457406997680664,
+      "step": 125
+    },
+    {
+      "epoch": 0.038300782734250326,
+      "grad_norm": 2.7533812522888184,
+      "learning_rate": 0.00015,
+      "loss": 5.223892688751221,
+      "step": 126
+    },
+    {
+      "epoch": 0.03860475720039517,
+      "grad_norm": 2.4422595500946045,
+      "learning_rate": 0.00015,
+      "loss": 5.512841701507568,
+      "step": 127
+    },
+    {
+      "epoch": 0.03890873166654001,
+      "grad_norm": 3.5897603034973145,
+      "learning_rate": 0.00015,
+      "loss": 4.829163551330566,
+      "step": 128
+    },
+    {
+      "epoch": 0.039212706132684856,
+      "grad_norm": 2.752082586288452,
+      "learning_rate": 0.00015,
+      "loss": 5.181515693664551,
+      "step": 129
+    },
+    {
+      "epoch": 0.0395166805988297,
+      "grad_norm": 2.35522723197937,
+      "learning_rate": 0.00015,
+      "loss": 4.889399528503418,
+      "step": 130
+    },
+    {
+      "epoch": 0.03982065506497454,
+      "grad_norm": 4.766695499420166,
+      "learning_rate": 0.00015,
+      "loss": 5.479020118713379,
+      "step": 131
+    },
+    {
+      "epoch": 0.040124629531119387,
+      "grad_norm": 3.1027352809906006,
+      "learning_rate": 0.00015,
+      "loss": 5.228668212890625,
+      "step": 132
+    },
+    {
+      "epoch": 0.04042860399726423,
+      "grad_norm": 2.8324410915374756,
+      "learning_rate": 0.00015,
+      "loss": 5.481418609619141,
+      "step": 133
+    },
+    {
+      "epoch": 0.04073257846340907,
+      "grad_norm": 3.3093934059143066,
+      "learning_rate": 0.00015,
+      "loss": 5.353525161743164,
+      "step": 134
+    },
+    {
+      "epoch": 0.04103655292955392,
+      "grad_norm": 2.960212469100952,
+      "learning_rate": 0.00015,
+      "loss": 5.206623554229736,
+      "step": 135
+    },
+    {
+      "epoch": 0.04134052739569876,
+      "grad_norm": 2.6950056552886963,
+      "learning_rate": 0.00015,
+      "loss": 5.788822174072266,
+      "step": 136
+    },
+    {
+      "epoch": 0.041644501861843604,
+      "grad_norm": 3.7148709297180176,
+      "learning_rate": 0.00015,
+      "loss": 5.405118465423584,
+      "step": 137
+    },
+    {
+      "epoch": 0.04194847632798845,
+      "grad_norm": 3.06892728805542,
+      "learning_rate": 0.00015,
+      "loss": 5.197515964508057,
+      "step": 138
+    },
+    {
+      "epoch": 0.04225245079413329,
+      "grad_norm": 2.5872280597686768,
+      "learning_rate": 0.00015,
+      "loss": 4.995965480804443,
+      "step": 139
+    },
+    {
+      "epoch": 0.042556425260278134,
+      "grad_norm": 2.83620285987854,
+      "learning_rate": 0.00015,
+      "loss": 5.071977138519287,
+      "step": 140
+    },
+    {
+      "epoch": 0.04286039972642298,
+      "grad_norm": 2.4734959602355957,
+      "learning_rate": 0.00015,
+      "loss": 5.329989433288574,
+      "step": 141
+    },
+    {
+      "epoch": 0.04316437419256782,
+      "grad_norm": 2.8380825519561768,
+      "learning_rate": 0.00015,
+      "loss": 5.345781326293945,
+      "step": 142
+    },
+    {
+      "epoch": 0.043468348658712665,
+      "grad_norm": 2.242863655090332,
+      "learning_rate": 0.00015,
+      "loss": 4.959817886352539,
+      "step": 143
+    },
+    {
+      "epoch": 0.043772323124857515,
+      "grad_norm": 2.8806676864624023,
+      "learning_rate": 0.00015,
+      "loss": 5.289271354675293,
+      "step": 144
+    },
+    {
+      "epoch": 0.04407629759100236,
+      "grad_norm": 2.490783929824829,
+      "learning_rate": 0.00015,
+      "loss": 5.227330207824707,
+      "step": 145
+    },
+    {
+      "epoch": 0.0443802720571472,
+      "grad_norm": 2.523484706878662,
+      "learning_rate": 0.00015,
+      "loss": 5.074975967407227,
+      "step": 146
+    },
+    {
+      "epoch": 0.044684246523292045,
+      "grad_norm": 3.964503288269043,
+      "learning_rate": 0.00015,
+      "loss": 5.131193161010742,
+      "step": 147
+    },
+    {
+      "epoch": 0.04498822098943689,
+      "grad_norm": 3.47658634185791,
+      "learning_rate": 0.00015,
+      "loss": 5.600862503051758,
+      "step": 148
+    },
+    {
+      "epoch": 0.04529219545558173,
+      "grad_norm": 3.7217631340026855,
+      "learning_rate": 0.00015,
+      "loss": 5.256523609161377,
+      "step": 149
+    },
+    {
+      "epoch": 0.045596169921726576,
+      "grad_norm": 3.3306832313537598,
+      "learning_rate": 0.00015,
+      "loss": 5.119356632232666,
+      "step": 150
+    },
+    {
+      "epoch": 0.04590014438787142,
+      "grad_norm": 3.020972728729248,
+      "learning_rate": 0.00015,
+      "loss": 5.328664779663086,
+      "step": 151
+    },
+    {
+      "epoch": 0.04620411885401626,
+      "grad_norm": 3.3983325958251953,
+      "learning_rate": 0.00015,
+      "loss": 5.074413776397705,
+      "step": 152
+    },
+    {
+      "epoch": 0.046508093320161106,
+      "grad_norm": 2.3080289363861084,
+      "learning_rate": 0.00015,
+      "loss": 5.054723739624023,
+      "step": 153
+    },
+    {
+      "epoch": 0.04681206778630595,
+      "grad_norm": 2.7934374809265137,
+      "learning_rate": 0.00015,
+      "loss": 5.081129550933838,
+      "step": 154
+    },
+    {
+      "epoch": 0.04711604225245079,
+      "grad_norm": 2.9598405361175537,
+      "learning_rate": 0.00015,
+      "loss": 5.2848100662231445,
+      "step": 155
+    },
+    {
+      "epoch": 0.04742001671859564,
+      "grad_norm": 2.4029855728149414,
+      "learning_rate": 0.00015,
+      "loss": 5.187831878662109,
+      "step": 156
+    },
+    {
+      "epoch": 0.04772399118474048,
+      "grad_norm": 2.6335527896881104,
+      "learning_rate": 0.00015,
+      "loss": 5.149965286254883,
+      "step": 157
+    },
+    {
+      "epoch": 0.048027965650885324,
+      "grad_norm": 4.03154993057251,
+      "learning_rate": 0.00015,
+      "loss": 5.30551815032959,
+      "step": 158
+    },
+    {
+      "epoch": 0.04833194011703017,
+      "grad_norm": 3.623112916946411,
+      "learning_rate": 0.00015,
+      "loss": 5.222558975219727,
+      "step": 159
+    },
+    {
+      "epoch": 0.04863591458317501,
+      "grad_norm": 2.9554989337921143,
+      "learning_rate": 0.00015,
+      "loss": 4.929510116577148,
+      "step": 160
+    },
+    {
+      "epoch": 0.048939889049319854,
+      "grad_norm": 3.150984048843384,
+      "learning_rate": 0.00015,
+      "loss": 5.1163740158081055,
+      "step": 161
+    },
+    {
+      "epoch": 0.0492438635154647,
+      "grad_norm": 2.9507365226745605,
+      "learning_rate": 0.00015,
+      "loss": 5.2223310470581055,
+      "step": 162
+    },
+    {
+      "epoch": 0.04954783798160955,
+      "grad_norm": 3.921109914779663,
+      "learning_rate": 0.00015,
+      "loss": 5.3069233894348145,
+      "step": 163
+    },
+    {
+      "epoch": 0.04985181244775439,
+      "grad_norm": 3.0457236766815186,
+      "learning_rate": 0.00015,
+      "loss": 5.484011650085449,
+      "step": 164
+    },
+    {
+      "epoch": 0.050155786913899235,
+      "grad_norm": 2.196756601333618,
+      "learning_rate": 0.00015,
+      "loss": 5.7509660720825195,
+      "step": 165
+    },
+    {
+      "epoch": 0.05045976138004408,
+      "grad_norm": 2.6914494037628174,
+      "learning_rate": 0.00015,
+      "loss": 5.604456424713135,
+      "step": 166
+    },
+    {
+      "epoch": 0.05076373584618892,
+      "grad_norm": 2.8087568283081055,
+      "learning_rate": 0.00015,
+      "loss": 5.164059162139893,
+      "step": 167
+    },
+    {
+      "epoch": 0.051067710312333765,
+      "grad_norm": 2.137946844100952,
+      "learning_rate": 0.00015,
+      "loss": 5.032990455627441,
+      "step": 168
+    },
+    {
+      "epoch": 0.05137168477847861,
+      "grad_norm": 2.417952299118042,
+      "learning_rate": 0.00015,
+      "loss": 5.011967658996582,
+      "step": 169
+    },
+    {
+      "epoch": 0.05167565924462345,
+      "grad_norm": 2.785905361175537,
+      "learning_rate": 0.00015,
+      "loss": 5.476579666137695,
+      "step": 170
+    },
+    {
+      "epoch": 0.051979633710768296,
+      "grad_norm": 2.5829813480377197,
+      "learning_rate": 0.00015,
+      "loss": 5.111513137817383,
+      "step": 171
+    },
+    {
+      "epoch": 0.05228360817691314,
+      "grad_norm": 1.989122986793518,
+      "learning_rate": 0.00015,
+      "loss": 5.1120710372924805,
+      "step": 172
+    },
+    {
+      "epoch": 0.05258758264305798,
+      "grad_norm": 2.241476058959961,
+      "learning_rate": 0.00015,
+      "loss": 5.431034088134766,
+      "step": 173
+    },
+    {
+      "epoch": 0.052891557109202826,
+      "grad_norm": 2.811575174331665,
+      "learning_rate": 0.00015,
+      "loss": 5.040464401245117,
+      "step": 174
+    },
+    {
+      "epoch": 0.05319553157534767,
+      "grad_norm": 2.502309799194336,
+      "learning_rate": 0.00015,
+      "loss": 5.1940693855285645,
+      "step": 175
+    },
+    {
+      "epoch": 0.05349950604149251,
+      "grad_norm": 2.620426654815674,
+      "learning_rate": 0.00015,
+      "loss": 5.009339809417725,
+      "step": 176
+    },
+    {
+      "epoch": 0.053803480507637357,
+      "grad_norm": 2.5785446166992188,
+      "learning_rate": 0.00015,
+      "loss": 5.791637420654297,
+      "step": 177
+    },
+    {
+      "epoch": 0.0541074549737822,
+      "grad_norm": 2.729437828063965,
+      "learning_rate": 0.00015,
+      "loss": 5.098715782165527,
+      "step": 178
+    },
+    {
+      "epoch": 0.05441142943992704,
+      "grad_norm": 2.4170680046081543,
+      "learning_rate": 0.00015,
+      "loss": 5.198185920715332,
+      "step": 179
+    },
+    {
+      "epoch": 0.05471540390607189,
+      "grad_norm": 3.367163896560669,
+      "learning_rate": 0.00015,
+      "loss": 5.431919097900391,
+      "step": 180
+    },
+    {
+      "epoch": 0.05501937837221673,
+      "grad_norm": 2.6199140548706055,
+      "learning_rate": 0.00015,
+      "loss": 5.356447696685791,
+      "step": 181
+    },
+    {
+      "epoch": 0.05532335283836158,
+      "grad_norm": 3.7180094718933105,
+      "learning_rate": 0.00015,
+      "loss": 5.236664772033691,
+      "step": 182
+    },
+    {
+      "epoch": 0.055627327304506424,
+      "grad_norm": 3.309356689453125,
+      "learning_rate": 0.00015,
+      "loss": 5.203059196472168,
+      "step": 183
+    },
+    {
+      "epoch": 0.05593130177065127,
+      "grad_norm": 2.2003111839294434,
+      "learning_rate": 0.00015,
+      "loss": 5.113945484161377,
+      "step": 184
+    },
+    {
+      "epoch": 0.05623527623679611,
+      "grad_norm": 2.6045517921447754,
+      "learning_rate": 0.00015,
+      "loss": 5.103667259216309,
+      "step": 185
+    },
+    {
+      "epoch": 0.056539250702940955,
+      "grad_norm": 3.046022891998291,
+      "learning_rate": 0.00015,
+      "loss": 5.524139404296875,
+      "step": 186
+    },
+    {
+      "epoch": 0.0568432251690858,
+      "grad_norm": 2.83263897895813,
+      "learning_rate": 0.00015,
+      "loss": 5.0012526512146,
+      "step": 187
+    },
+    {
+      "epoch": 0.05714719963523064,
+      "grad_norm": 2.2974939346313477,
+      "learning_rate": 0.00015,
+      "loss": 5.0340800285339355,
+      "step": 188
+    },
+    {
+      "epoch": 0.057451174101375485,
+      "grad_norm": 2.611394166946411,
+      "learning_rate": 0.00015,
+      "loss": 5.159695625305176,
+      "step": 189
+    },
+    {
+      "epoch": 0.05775514856752033,
+      "grad_norm": 2.625948429107666,
+      "learning_rate": 0.00015,
+      "loss": 5.3006696701049805,
+      "step": 190
+    },
+    {
+      "epoch": 0.05805912303366517,
+      "grad_norm": 3.0314557552337646,
+      "learning_rate": 0.00015,
+      "loss": 5.176998138427734,
+      "step": 191
+    },
+    {
+      "epoch": 0.058363097499810015,
+      "grad_norm": 2.3988449573516846,
+      "learning_rate": 0.00015,
+      "loss": 5.312588691711426,
+      "step": 192
+    },
+    {
+      "epoch": 0.05866707196595486,
+      "grad_norm": 2.5515637397766113,
+      "learning_rate": 0.00015,
+      "loss": 4.93233585357666,
+      "step": 193
+    },
+    {
+      "epoch": 0.0589710464320997,
+      "grad_norm": 3.2827372550964355,
+      "learning_rate": 0.00015,
+      "loss": 4.914484977722168,
+      "step": 194
+    },
+    {
+      "epoch": 0.059275020898244546,
+      "grad_norm": 2.499932050704956,
+      "learning_rate": 0.00015,
+      "loss": 5.154760360717773,
+      "step": 195
+    },
+    {
+      "epoch": 0.05957899536438939,
+      "grad_norm": 2.637345314025879,
+      "learning_rate": 0.00015,
+      "loss": 5.163487434387207,
+      "step": 196
+    },
+    {
+      "epoch": 0.05988296983053423,
+      "grad_norm": 2.3949882984161377,
+      "learning_rate": 0.00015,
+      "loss": 5.231352806091309,
+      "step": 197
+    },
+    {
+      "epoch": 0.060186944296679076,
+      "grad_norm": 4.018889427185059,
+      "learning_rate": 0.00015,
+      "loss": 5.185758590698242,
+      "step": 198
+    },
+    {
+      "epoch": 0.06049091876282392,
+      "grad_norm": 2.689361333847046,
+      "learning_rate": 0.00015,
+      "loss": 5.079290866851807,
+      "step": 199
+    },
+    {
+      "epoch": 0.06079489322896876,
+      "grad_norm": 2.546954393386841,
+      "learning_rate": 0.00015,
+      "loss": 5.032594680786133,
+      "step": 200
+    },
+    {
+      "epoch": 0.061098867695113614,
+      "grad_norm": 2.464322328567505,
+      "learning_rate": 0.00015,
+      "loss": 5.091517448425293,
+      "step": 201
+    },
+    {
+      "epoch": 0.06140284216125846,
+      "grad_norm": 2.063349723815918,
+      "learning_rate": 0.00015,
+      "loss": 4.76100492477417,
+      "step": 202
+    },
+    {
+      "epoch": 0.0617068166274033,
+      "grad_norm": 2.0247418880462646,
+      "learning_rate": 0.00015,
+      "loss": 4.9118428230285645,
+      "step": 203
+    },
+    {
+      "epoch": 0.062010791093548144,
+      "grad_norm": 2.9584648609161377,
+      "learning_rate": 0.00015,
+      "loss": 5.26734733581543,
+      "step": 204
+    },
+    {
+      "epoch": 0.06231476555969299,
+      "grad_norm": 2.1421217918395996,
+      "learning_rate": 0.00015,
+      "loss": 5.158268451690674,
+      "step": 205
+    },
+    {
+      "epoch": 0.06261874002583782,
+      "grad_norm": 2.246140241622925,
+      "learning_rate": 0.00015,
+      "loss": 4.949875831604004,
+      "step": 206
+    },
+    {
+      "epoch": 0.06292271449198267,
+      "grad_norm": 2.209535837173462,
+      "learning_rate": 0.00015,
+      "loss": 4.945962429046631,
+      "step": 207
+    },
+    {
+      "epoch": 0.06322668895812751,
+      "grad_norm": 2.561234951019287,
+      "learning_rate": 0.00015,
+      "loss": 5.304056644439697,
+      "step": 208
+    },
+    {
+      "epoch": 0.06353066342427235,
+      "grad_norm": 2.179927110671997,
+      "learning_rate": 0.00015,
+      "loss": 4.938395977020264,
+      "step": 209
+    },
+    {
+      "epoch": 0.06383463789041721,
+      "grad_norm": 2.645345449447632,
+      "learning_rate": 0.00015,
+      "loss": 4.911320686340332,
+      "step": 210
+    },
+    {
+      "epoch": 0.06413861235656206,
+      "grad_norm": 2.700261116027832,
+      "learning_rate": 0.00015,
+      "loss": 5.000833034515381,
+      "step": 211
+    },
+    {
+      "epoch": 0.0644425868227069,
+      "grad_norm": 1.9913718700408936,
+      "learning_rate": 0.00015,
+      "loss": 5.216194152832031,
+      "step": 212
+    },
+    {
+      "epoch": 0.06474656128885174,
+      "grad_norm": 5.805688858032227,
+      "learning_rate": 0.00015,
+      "loss": 5.376114845275879,
+      "step": 213
+    },
+    {
+      "epoch": 0.06505053575499659,
+      "grad_norm": 2.8809239864349365,
+      "learning_rate": 0.00015,
+      "loss": 5.206188201904297,
+      "step": 214
+    },
+    {
+      "epoch": 0.06535451022114143,
+      "grad_norm": 2.6883764266967773,
+      "learning_rate": 0.00015,
+      "loss": 5.38466739654541,
+      "step": 215
+    },
+    {
+      "epoch": 0.06565848468728627,
+      "grad_norm": 2.3931589126586914,
+      "learning_rate": 0.00015,
+      "loss": 5.033239364624023,
+      "step": 216
+    },
+    {
+      "epoch": 0.06596245915343112,
+      "grad_norm": 2.618607759475708,
+      "learning_rate": 0.00015,
+      "loss": 5.279989242553711,
+      "step": 217
+    },
+    {
+      "epoch": 0.06626643361957596,
+      "grad_norm": 2.409651041030884,
+      "learning_rate": 0.00015,
+      "loss": 4.695886135101318,
+      "step": 218
+    },
+    {
+      "epoch": 0.0665704080857208,
+      "grad_norm": 3.0053608417510986,
+      "learning_rate": 0.00015,
+      "loss": 5.499357223510742,
+      "step": 219
+    },
+    {
+      "epoch": 0.06687438255186565,
+      "grad_norm": 2.2576591968536377,
+      "learning_rate": 0.00015,
+      "loss": 4.99501895904541,
+      "step": 220
+    },
+    {
+      "epoch": 0.06717835701801049,
+      "grad_norm": 2.346895217895508,
+      "learning_rate": 0.00015,
+      "loss": 5.231335639953613,
+      "step": 221
+    },
+    {
+      "epoch": 0.06748233148415533,
+      "grad_norm": 2.6597862243652344,
+      "learning_rate": 0.00015,
+      "loss": 5.022856712341309,
+      "step": 222
+    },
+    {
+      "epoch": 0.06778630595030018,
+      "grad_norm": 2.602850914001465,
+      "learning_rate": 0.00015,
+      "loss": 5.263967990875244,
+      "step": 223
+    },
+    {
+      "epoch": 0.06809028041644502,
+      "grad_norm": 2.488557815551758,
+      "learning_rate": 0.00015,
+      "loss": 5.073237895965576,
+      "step": 224
+    },
+    {
+      "epoch": 0.06839425488258986,
+      "grad_norm": 3.207136392593384,
+      "learning_rate": 0.00015,
+      "loss": 5.085233211517334,
+      "step": 225
+    },
+    {
+      "epoch": 0.06869822934873471,
+      "grad_norm": 2.5358049869537354,
+      "learning_rate": 0.00015,
+      "loss": 5.051425933837891,
+      "step": 226
+    },
+    {
+      "epoch": 0.06900220381487955,
+      "grad_norm": 2.7574715614318848,
+      "learning_rate": 0.00015,
+      "loss": 4.930265426635742,
+      "step": 227
+    },
+    {
+      "epoch": 0.0693061782810244,
+      "grad_norm": 2.285963535308838,
+      "learning_rate": 0.00015,
+      "loss": 5.425419807434082,
+      "step": 228
+    },
+    {
+      "epoch": 0.06961015274716924,
+      "grad_norm": 2.759037733078003,
+      "learning_rate": 0.00015,
+      "loss": 4.854893684387207,
+      "step": 229
+    },
+    {
+      "epoch": 0.06991412721331408,
+      "grad_norm": 2.5253970623016357,
+      "learning_rate": 0.00015,
+      "loss": 4.819305896759033,
+      "step": 230
+    },
+    {
+      "epoch": 0.07021810167945892,
+      "grad_norm": 3.8112499713897705,
+      "learning_rate": 0.00015,
+      "loss": 4.986693382263184,
+      "step": 231
+    },
+    {
+      "epoch": 0.07052207614560377,
+      "grad_norm": 2.5096371173858643,
+      "learning_rate": 0.00015,
+      "loss": 5.021882057189941,
+      "step": 232
+    },
+    {
+      "epoch": 0.07082605061174861,
+      "grad_norm": 4.503645420074463,
+      "learning_rate": 0.00015,
+      "loss": 4.774106502532959,
+      "step": 233
+    },
+    {
+      "epoch": 0.07113002507789346,
+      "grad_norm": 3.059335708618164,
+      "learning_rate": 0.00015,
+      "loss": 4.8400115966796875,
+      "step": 234
+    },
+    {
+      "epoch": 0.0714339995440383,
+      "grad_norm": 2.524287223815918,
+      "learning_rate": 0.00015,
+      "loss": 5.0765533447265625,
+      "step": 235
+    },
+    {
+      "epoch": 0.07173797401018314,
+      "grad_norm": 2.644411087036133,
+      "learning_rate": 0.00015,
+      "loss": 5.126844882965088,
+      "step": 236
+    },
+    {
+      "epoch": 0.07204194847632799,
+      "grad_norm": 2.767155408859253,
+      "learning_rate": 0.00015,
+      "loss": 5.2514424324035645,
+      "step": 237
+    },
+    {
+      "epoch": 0.07234592294247283,
+      "grad_norm": 2.7995824813842773,
+      "learning_rate": 0.00015,
+      "loss": 5.195583343505859,
+      "step": 238
+    },
+    {
+      "epoch": 0.07264989740861767,
+      "grad_norm": 1.8425229787826538,
+      "learning_rate": 0.00015,
+      "loss": 4.947050094604492,
+      "step": 239
+    },
+    {
+      "epoch": 0.07295387187476252,
+      "grad_norm": 2.1151154041290283,
+      "learning_rate": 0.00015,
+      "loss": 4.862464904785156,
+      "step": 240
+    },
+    {
+      "epoch": 0.07325784634090736,
+      "grad_norm": 1.8664017915725708,
+      "learning_rate": 0.00015,
+      "loss": 5.00832462310791,
+      "step": 241
+    },
+    {
+      "epoch": 0.0735618208070522,
+      "grad_norm": 2.731326103210449,
+      "learning_rate": 0.00015,
+      "loss": 5.133040428161621,
+      "step": 242
+    },
+    {
+      "epoch": 0.07386579527319705,
+      "grad_norm": 2.1356990337371826,
+      "learning_rate": 0.00015,
+      "loss": 4.863617420196533,
+      "step": 243
+    },
+    {
+      "epoch": 0.07416976973934189,
+      "grad_norm": 2.531816005706787,
+      "learning_rate": 0.00015,
+      "loss": 4.960110664367676,
+      "step": 244
+    },
+    {
+      "epoch": 0.07447374420548673,
+      "grad_norm": 2.294783592224121,
+      "learning_rate": 0.00015,
+      "loss": 5.038862228393555,
+      "step": 245
+    },
+    {
+      "epoch": 0.07477771867163158,
+      "grad_norm": 2.583592176437378,
+      "learning_rate": 0.00015,
+      "loss": 4.6658101081848145,
+      "step": 246
+    },
+    {
+      "epoch": 0.07508169313777642,
+      "grad_norm": 2.7309248447418213,
+      "learning_rate": 0.00015,
+      "loss": 5.1576151847839355,
+      "step": 247
+    },
+    {
+      "epoch": 0.07538566760392128,
+      "grad_norm": 2.7108030319213867,
+      "learning_rate": 0.00015,
+      "loss": 4.991024017333984,
+      "step": 248
+    },
+    {
+      "epoch": 0.07568964207006612,
+      "grad_norm": 2.1597740650177,
+      "learning_rate": 0.00015,
+      "loss": 5.090862274169922,
+      "step": 249
+    },
+    {
+      "epoch": 0.07599361653621096,
+      "grad_norm": 3.0891337394714355,
+      "learning_rate": 0.00015,
+      "loss": 5.252491474151611,
+      "step": 250
+    },
+    {
+      "epoch": 0.07629759100235581,
+      "grad_norm": 2.791867256164551,
+      "learning_rate": 0.00015,
+      "loss": 4.918403148651123,
+      "step": 251
+    },
+    {
+      "epoch": 0.07660156546850065,
+      "grad_norm": 4.059783935546875,
+      "learning_rate": 0.00015,
+      "loss": 5.10411262512207,
+      "step": 252
+    },
+    {
+      "epoch": 0.0769055399346455,
+      "grad_norm": 3.711529016494751,
+      "learning_rate": 0.00015,
+      "loss": 5.02101469039917,
+      "step": 253
+    },
+    {
+      "epoch": 0.07720951440079034,
+      "grad_norm": 2.7228102684020996,
+      "learning_rate": 0.00015,
+      "loss": 5.0215840339660645,
+      "step": 254
+    },
+    {
+      "epoch": 0.07751348886693518,
+      "grad_norm": 2.5281450748443604,
+      "learning_rate": 0.00015,
+      "loss": 4.998557090759277,
+      "step": 255
+    },
+    {
+      "epoch": 0.07781746333308003,
+      "grad_norm": 2.29402494430542,
+      "learning_rate": 0.00015,
+      "loss": 5.168207168579102,
+      "step": 256
+    },
+    {
+      "epoch": 0.07812143779922487,
+      "grad_norm": 2.665695905685425,
+      "learning_rate": 0.00015,
+      "loss": 5.190396308898926,
+      "step": 257
+    },
+    {
+      "epoch": 0.07842541226536971,
+      "grad_norm": 2.947962522506714,
+      "learning_rate": 0.00015,
+      "loss": 5.241790771484375,
+      "step": 258
+    },
+    {
+      "epoch": 0.07872938673151456,
+      "grad_norm": 3.139033079147339,
+      "learning_rate": 0.00015,
+      "loss": 5.01392126083374,
+      "step": 259
+    },
+    {
+      "epoch": 0.0790333611976594,
+      "grad_norm": 2.243774175643921,
+      "learning_rate": 0.00015,
+      "loss": 4.711472511291504,
+      "step": 260
+    },
+    {
+      "epoch": 0.07933733566380424,
+      "grad_norm": 2.432201862335205,
+      "learning_rate": 0.00015,
+      "loss": 4.898543834686279,
+      "step": 261
+    },
+    {
+      "epoch": 0.07964131012994909,
+      "grad_norm": 2.25750994682312,
+      "learning_rate": 0.00015,
+      "loss": 4.939959526062012,
+      "step": 262
+    },
+    {
+      "epoch": 0.07994528459609393,
+      "grad_norm": 2.3874895572662354,
+      "learning_rate": 0.00015,
+      "loss": 4.840463638305664,
+      "step": 263
+    },
+    {
+      "epoch": 0.08024925906223877,
+      "grad_norm": 2.496791124343872,
+      "learning_rate": 0.00015,
+      "loss": 5.27952241897583,
+      "step": 264
+    },
+    {
+      "epoch": 0.08055323352838362,
+      "grad_norm": 2.284442901611328,
+      "learning_rate": 0.00015,
+      "loss": 4.927646636962891,
+      "step": 265
+    },
+    {
+      "epoch": 0.08085720799452846,
+      "grad_norm": 3.212139844894409,
+      "learning_rate": 0.00015,
+      "loss": 5.056288719177246,
+      "step": 266
+    },
+    {
+      "epoch": 0.0811611824606733,
+      "grad_norm": 2.5842652320861816,
+      "learning_rate": 0.00015,
+      "loss": 5.33665657043457,
+      "step": 267
+    },
+    {
+      "epoch": 0.08146515692681815,
+      "grad_norm": 2.2716422080993652,
+      "learning_rate": 0.00015,
+      "loss": 4.699724197387695,
+      "step": 268
+    },
+    {
+      "epoch": 0.08176913139296299,
+      "grad_norm": 2.4571189880371094,
+      "learning_rate": 0.00015,
+      "loss": 5.148621082305908,
+      "step": 269
+    },
+    {
+      "epoch": 0.08207310585910783,
+      "grad_norm": 2.7588019371032715,
+      "learning_rate": 0.00015,
+      "loss": 5.022087574005127,
+      "step": 270
+    },
+    {
+      "epoch": 0.08237708032525268,
+      "grad_norm": 2.6230340003967285,
+      "learning_rate": 0.00015,
+      "loss": 5.289278507232666,
+      "step": 271
+    },
+    {
+      "epoch": 0.08268105479139752,
+      "grad_norm": 2.8345372676849365,
+      "learning_rate": 0.00015,
+      "loss": 5.0885138511657715,
+      "step": 272
+    },
+    {
+      "epoch": 0.08298502925754236,
+      "grad_norm": 2.9786059856414795,
+      "learning_rate": 0.00015,
+      "loss": 4.860764026641846,
+      "step": 273
+    },
+    {
+      "epoch": 0.08328900372368721,
+      "grad_norm": 2.058086395263672,
+      "learning_rate": 0.00015,
+      "loss": 4.823609352111816,
+      "step": 274
+    },
+    {
+      "epoch": 0.08359297818983205,
+      "grad_norm": 2.5309677124023438,
+      "learning_rate": 0.00015,
+      "loss": 5.097412109375,
+      "step": 275
+    },
+    {
+      "epoch": 0.0838969526559769,
+      "grad_norm": 2.7556543350219727,
+      "learning_rate": 0.00015,
+      "loss": 5.104552268981934,
+      "step": 276
+    },
+    {
+      "epoch": 0.08420092712212174,
+      "grad_norm": 2.7713849544525146,
+      "learning_rate": 0.00015,
+      "loss": 4.843084812164307,
+      "step": 277
+    },
+    {
+      "epoch": 0.08450490158826658,
+      "grad_norm": 2.2433159351348877,
+      "learning_rate": 0.00015,
+      "loss": 4.960465431213379,
+      "step": 278
+    },
+    {
+      "epoch": 0.08480887605441143,
+      "grad_norm": 2.418747901916504,
+      "learning_rate": 0.00015,
+      "loss": 4.657155990600586,
+      "step": 279
+    },
+    {
+      "epoch": 0.08511285052055627,
+      "grad_norm": 4.219315528869629,
+      "learning_rate": 0.00015,
+      "loss": 5.192144393920898,
+      "step": 280
+    },
+    {
+      "epoch": 0.08541682498670111,
+      "grad_norm": 3.733152151107788,
+      "learning_rate": 0.00015,
+      "loss": 4.924145698547363,
+      "step": 281
+    },
+    {
+      "epoch": 0.08572079945284596,
+      "grad_norm": 2.8210558891296387,
+      "learning_rate": 0.00015,
+      "loss": 5.179683685302734,
+      "step": 282
+    },
+    {
+      "epoch": 0.0860247739189908,
+      "grad_norm": 2.3206191062927246,
+      "learning_rate": 0.00015,
+      "loss": 5.157569408416748,
+      "step": 283
+    },
+    {
+      "epoch": 0.08632874838513564,
+      "grad_norm": 2.8121376037597656,
+      "learning_rate": 0.00015,
+      "loss": 4.984990119934082,
+      "step": 284
+    },
+    {
+      "epoch": 0.08663272285128049,
+      "grad_norm": 2.259294033050537,
+      "learning_rate": 0.00015,
+      "loss": 5.096368312835693,
+      "step": 285
+    },
+    {
+      "epoch": 0.08693669731742533,
+      "grad_norm": 2.4496352672576904,
+      "learning_rate": 0.00015,
+      "loss": 4.88273811340332,
+      "step": 286
+    },
+    {
+      "epoch": 0.08724067178357019,
+      "grad_norm": 2.86950945854187,
+      "learning_rate": 0.00015,
+      "loss": 4.964693069458008,
+      "step": 287
+    },
+    {
+      "epoch": 0.08754464624971503,
+      "grad_norm": 2.428152322769165,
+      "learning_rate": 0.00015,
+      "loss": 5.048365592956543,
+      "step": 288
+    },
+    {
+      "epoch": 0.08784862071585987,
+      "grad_norm": 2.44697642326355,
+      "learning_rate": 0.00015,
+      "loss": 5.177502632141113,
+      "step": 289
+    },
+    {
+      "epoch": 0.08815259518200472,
+      "grad_norm": 2.8711910247802734,
+      "learning_rate": 0.00015,
+      "loss": 4.932013511657715,
+      "step": 290
+    },
+    {
+      "epoch": 0.08845656964814956,
+      "grad_norm": 2.2881672382354736,
+      "learning_rate": 0.00015,
+      "loss": 4.936160087585449,
+      "step": 291
+    },
+    {
+      "epoch": 0.0887605441142944,
+      "grad_norm": 2.057288885116577,
+      "learning_rate": 0.00015,
+      "loss": 4.9758195877075195,
+      "step": 292
+    },
+    {
+      "epoch": 0.08906451858043925,
+      "grad_norm": 2.7649283409118652,
+      "learning_rate": 0.00015,
+      "loss": 4.989021301269531,
+      "step": 293
+    },
+    {
+      "epoch": 0.08936849304658409,
+      "grad_norm": 2.789336919784546,
+      "learning_rate": 0.00015,
+      "loss": 5.080655097961426,
+      "step": 294
+    },
+    {
+      "epoch": 0.08967246751272893,
+      "grad_norm": 2.9518494606018066,
+      "learning_rate": 0.00015,
+      "loss": 4.984024524688721,
+      "step": 295
+    },
+    {
+      "epoch": 0.08997644197887378,
+      "grad_norm": 2.447422504425049,
+      "learning_rate": 0.00015,
+      "loss": 5.239782333374023,
+      "step": 296
+    },
+    {
+      "epoch": 0.09028041644501862,
+      "grad_norm": 2.062412738800049,
+      "learning_rate": 0.00015,
+      "loss": 5.112987518310547,
+      "step": 297
+    },
+    {
+      "epoch": 0.09058439091116346,
+      "grad_norm": 2.1599228382110596,
+      "learning_rate": 0.00015,
+      "loss": 5.321422576904297,
+      "step": 298
+    },
+    {
+      "epoch": 0.09088836537730831,
+      "grad_norm": 2.3566126823425293,
+      "learning_rate": 0.00015,
+      "loss": 4.953940391540527,
+      "step": 299
+    },
+    {
+      "epoch": 0.09119233984345315,
+      "grad_norm": 2.4276676177978516,
+      "learning_rate": 0.00015,
+      "loss": 4.933962821960449,
+      "step": 300
+    },
+    {
+      "epoch": 0.091496314309598,
+      "grad_norm": 2.4276134967803955,
+      "learning_rate": 0.00015,
+      "loss": 5.000826835632324,
+      "step": 301
+    },
+    {
+      "epoch": 0.09180028877574284,
+      "grad_norm": 2.401012897491455,
+      "learning_rate": 0.00015,
+      "loss": 5.0254411697387695,
+      "step": 302
+    },
+    {
+      "epoch": 0.09210426324188768,
+      "grad_norm": 2.1861603260040283,
+      "learning_rate": 0.00015,
+      "loss": 4.874423027038574,
+      "step": 303
+    },
+    {
+      "epoch": 0.09240823770803253,
+      "grad_norm": 2.7462263107299805,
+      "learning_rate": 0.00015,
+      "loss": 4.771231651306152,
+      "step": 304
+    },
+    {
+      "epoch": 0.09271221217417737,
+      "grad_norm": 2.331862449645996,
+      "learning_rate": 0.00015,
+      "loss": 4.694076061248779,
+      "step": 305
+    },
+    {
+      "epoch": 0.09301618664032221,
+      "grad_norm": 2.3344733715057373,
+      "learning_rate": 0.00015,
+      "loss": 4.892954349517822,
+      "step": 306
+    },
+    {
+      "epoch": 0.09332016110646706,
+      "grad_norm": 2.705339193344116,
+      "learning_rate": 0.00015,
+      "loss": 4.8948798179626465,
+      "step": 307
+    },
+    {
+      "epoch": 0.0936241355726119,
+      "grad_norm": 2.401493787765503,
+      "learning_rate": 0.00015,
+      "loss": 4.859251022338867,
+      "step": 308
+    },
+    {
+      "epoch": 0.09392811003875674,
+      "grad_norm": 2.151730537414551,
+      "learning_rate": 0.00015,
+      "loss": 4.9333343505859375,
+      "step": 309
+    },
+    {
+      "epoch": 0.09423208450490159,
+      "grad_norm": 2.1982614994049072,
+      "learning_rate": 0.00015,
+      "loss": 5.303628921508789,
+      "step": 310
+    },
+    {
+      "epoch": 0.09453605897104643,
+      "grad_norm": 2.652592420578003,
+      "learning_rate": 0.00015,
+      "loss": 4.941991329193115,
+      "step": 311
+    },
+    {
+      "epoch": 0.09484003343719127,
+      "grad_norm": 2.4708220958709717,
+      "learning_rate": 0.00015,
+      "loss": 5.204559326171875,
+      "step": 312
+    },
+    {
+      "epoch": 0.09514400790333612,
+      "grad_norm": 2.733691930770874,
+      "learning_rate": 0.00015,
+      "loss": 5.362415790557861,
+      "step": 313
+    },
+    {
+      "epoch": 0.09544798236948096,
+      "grad_norm": 2.116511106491089,
+      "learning_rate": 0.00015,
+      "loss": 4.778589248657227,
+      "step": 314
+    },
+    {
+      "epoch": 0.0957519568356258,
+      "grad_norm": 2.324578285217285,
+      "learning_rate": 0.00015,
+      "loss": 5.327642917633057,
+      "step": 315
+    },
+    {
+      "epoch": 0.09605593130177065,
+      "grad_norm": 2.1824278831481934,
+      "learning_rate": 0.00015,
+      "loss": 5.259299278259277,
+      "step": 316
+    },
+    {
+      "epoch": 0.09635990576791549,
+      "grad_norm": 2.7062108516693115,
+      "learning_rate": 0.00015,
+      "loss": 5.072212219238281,
+      "step": 317
+    },
+    {
+      "epoch": 0.09666388023406033,
+      "grad_norm": 2.5505473613739014,
+      "learning_rate": 0.00015,
+      "loss": 4.9667205810546875,
+      "step": 318
+    },
+    {
+      "epoch": 0.09696785470020518,
+      "grad_norm": 2.4435019493103027,
+      "learning_rate": 0.00015,
+      "loss": 5.01628303527832,
+      "step": 319
+    },
+    {
+      "epoch": 0.09727182916635002,
+      "grad_norm": 2.3002214431762695,
+      "learning_rate": 0.00015,
+      "loss": 4.931803226470947,
+      "step": 320
+    },
+    {
+      "epoch": 0.09757580363249486,
+      "grad_norm": 2.0840365886688232,
+      "learning_rate": 0.00015,
+      "loss": 5.155757904052734,
+      "step": 321
+    },
+    {
+      "epoch": 0.09787977809863971,
+      "grad_norm": 2.4219934940338135,
+      "learning_rate": 0.00015,
+      "loss": 5.274277210235596,
+      "step": 322
+    },
+    {
+      "epoch": 0.09818375256478455,
+      "grad_norm": 2.458807945251465,
+      "learning_rate": 0.00015,
+      "loss": 4.875680923461914,
+      "step": 323
+    },
+    {
+      "epoch": 0.0984877270309294,
+      "grad_norm": 2.6304242610931396,
+      "learning_rate": 0.00015,
+      "loss": 5.189663887023926,
+      "step": 324
+    },
+    {
+      "epoch": 0.09879170149707425,
+      "grad_norm": 2.5474510192871094,
+      "learning_rate": 0.00015,
+      "loss": 4.984101295471191,
+      "step": 325
+    },
+    {
+      "epoch": 0.0990956759632191,
+      "grad_norm": 2.3748531341552734,
+      "learning_rate": 0.00015,
+      "loss": 5.225240230560303,
+      "step": 326
+    },
+    {
+      "epoch": 0.09939965042936394,
+      "grad_norm": 2.3719253540039062,
+      "learning_rate": 0.00015,
+      "loss": 4.587345123291016,
+      "step": 327
+    },
+    {
+      "epoch": 0.09970362489550878,
+      "grad_norm": 2.314894676208496,
+      "learning_rate": 0.00015,
+      "loss": 4.955268383026123,
+      "step": 328
+    },
+    {
+      "epoch": 0.10000759936165363,
+      "grad_norm": 2.4700207710266113,
+      "learning_rate": 0.00015,
+      "loss": 5.03272008895874,
+      "step": 329
+    },
+    {
+      "epoch": 0.10031157382779847,
+      "grad_norm": 2.0580408573150635,
+      "learning_rate": 0.00015,
+      "loss": 5.035648822784424,
+      "step": 330
+    },
+    {
+      "epoch": 0.10061554829394331,
+      "grad_norm": 2.475537061691284,
+      "learning_rate": 0.00015,
+      "loss": 5.316486358642578,
+      "step": 331
+    },
+    {
+      "epoch": 0.10091952276008816,
+      "grad_norm": 2.1298344135284424,
+      "learning_rate": 0.00015,
+      "loss": 4.8759965896606445,
+      "step": 332
+    },
+    {
+      "epoch": 0.101223497226233,
+      "grad_norm": 3.1105642318725586,
+      "learning_rate": 0.00015,
+      "loss": 5.251327991485596,
+      "step": 333
+    },
+    {
+      "epoch": 0.10152747169237784,
+      "grad_norm": 2.6745150089263916,
+      "learning_rate": 0.00015,
+      "loss": 5.150968074798584,
+      "step": 334
+    },
+    {
+      "epoch": 0.10183144615852269,
+      "grad_norm": 2.207319498062134,
+      "learning_rate": 0.00015,
+      "loss": 4.971871852874756,
+      "step": 335
+    },
+    {
+      "epoch": 0.10213542062466753,
+      "grad_norm": 2.583159923553467,
+      "learning_rate": 0.00015,
+      "loss": 4.795596599578857,
+      "step": 336
+    },
+    {
+      "epoch": 0.10243939509081237,
+      "grad_norm": 2.4274961948394775,
+      "learning_rate": 0.00015,
+      "loss": 4.810564994812012,
+      "step": 337
+    },
+    {
+      "epoch": 0.10274336955695722,
+      "grad_norm": 2.5559823513031006,
+      "learning_rate": 0.00015,
+      "loss": 5.243542671203613,
+      "step": 338
+    },
+    {
+      "epoch": 0.10304734402310206,
+      "grad_norm": 2.270174264907837,
+      "learning_rate": 0.00015,
+      "loss": 4.670610427856445,
+      "step": 339
+    },
+    {
+      "epoch": 0.1033513184892469,
+      "grad_norm": 2.6501379013061523,
+      "learning_rate": 0.00015,
+      "loss": 4.815388202667236,
+      "step": 340
+    },
+    {
+      "epoch": 0.10365529295539175,
+      "grad_norm": 2.0966672897338867,
+      "learning_rate": 0.00015,
+      "loss": 4.696559429168701,
+      "step": 341
+    },
+    {
+      "epoch": 0.10395926742153659,
+      "grad_norm": 3.616415023803711,
+      "learning_rate": 0.00015,
+      "loss": 5.159313201904297,
+      "step": 342
+    },
+    {
+      "epoch": 0.10426324188768143,
+      "grad_norm": 2.887364149093628,
+      "learning_rate": 0.00015,
+      "loss": 5.051226615905762,
+      "step": 343
+    },
+    {
+      "epoch": 0.10456721635382628,
+      "grad_norm": 2.8242428302764893,
+      "learning_rate": 0.00015,
+      "loss": 4.752954006195068,
+      "step": 344
+    },
+    {
+      "epoch": 0.10487119081997112,
+      "grad_norm": 2.3553125858306885,
+      "learning_rate": 0.00015,
+      "loss": 4.793919563293457,
+      "step": 345
+    },
+    {
+      "epoch": 0.10517516528611597,
+      "grad_norm": 3.5092177391052246,
+      "learning_rate": 0.00015,
+      "loss": 4.906852722167969,
+      "step": 346
+    },
+    {
+      "epoch": 0.10547913975226081,
+      "grad_norm": 2.2616965770721436,
+      "learning_rate": 0.00015,
+      "loss": 5.071382522583008,
+      "step": 347
+    },
+    {
+      "epoch": 0.10578311421840565,
+      "grad_norm": 2.088045835494995,
+      "learning_rate": 0.00015,
+      "loss": 5.169279098510742,
+      "step": 348
+    },
+    {
+      "epoch": 0.1060870886845505,
+      "grad_norm": 2.0404839515686035,
+      "learning_rate": 0.00015,
+      "loss": 5.174063205718994,
+      "step": 349
+    },
+    {
+      "epoch": 0.10639106315069534,
+      "grad_norm": 3.254695415496826,
+      "learning_rate": 0.00015,
+      "loss": 4.8464202880859375,
+      "step": 350
+    },
+    {
+      "epoch": 0.10669503761684018,
+      "grad_norm": 2.800034284591675,
+      "learning_rate": 0.00015,
+      "loss": 4.866253852844238,
+      "step": 351
+    },
+    {
+      "epoch": 0.10699901208298503,
+      "grad_norm": 2.4757239818573,
+      "learning_rate": 0.00015,
+      "loss": 4.849142551422119,
+      "step": 352
+    },
+    {
+      "epoch": 0.10730298654912987,
+      "grad_norm": 2.6723878383636475,
+      "learning_rate": 0.00015,
+      "loss": 5.3866987228393555,
+      "step": 353
+    },
+    {
+      "epoch": 0.10760696101527471,
+      "grad_norm": 2.601670026779175,
+      "learning_rate": 0.00015,
+      "loss": 4.975025177001953,
+      "step": 354
+    },
+    {
+      "epoch": 0.10791093548141956,
+      "grad_norm": 2.8854212760925293,
+      "learning_rate": 0.00015,
+      "loss": 4.811579704284668,
+      "step": 355
+    },
+    {
+      "epoch": 0.1082149099475644,
+      "grad_norm": 3.1337034702301025,
+      "learning_rate": 0.00015,
+      "loss": 4.607230186462402,
+      "step": 356
+    },
+    {
+      "epoch": 0.10851888441370924,
+      "grad_norm": 2.2333157062530518,
+      "learning_rate": 0.00015,
+      "loss": 4.906778335571289,
+      "step": 357
+    },
+    {
+      "epoch": 0.10882285887985409,
+      "grad_norm": 2.1132254600524902,
+      "learning_rate": 0.00015,
+      "loss": 4.8923020362854,
+      "step": 358
+    },
+    {
+      "epoch": 0.10912683334599893,
+      "grad_norm": 2.108551263809204,
+      "learning_rate": 0.00015,
+      "loss": 4.993260383605957,
+      "step": 359
+    },
+    {
+      "epoch": 0.10943080781214377,
+      "grad_norm": 2.6499183177948,
+      "learning_rate": 0.00015,
+      "loss": 5.038416862487793,
+      "step": 360
+    },
+    {
+      "epoch": 0.10973478227828862,
+      "grad_norm": 2.1935219764709473,
+      "learning_rate": 0.00015,
+      "loss": 5.338360786437988,
+      "step": 361
+    },
+    {
+      "epoch": 0.11003875674443346,
+      "grad_norm": 2.966428518295288,
+      "learning_rate": 0.00015,
+      "loss": 5.086145877838135,
+      "step": 362
+    },
+    {
+      "epoch": 0.11034273121057832,
+      "grad_norm": 2.0816824436187744,
+      "learning_rate": 0.00015,
+      "loss": 4.8238301277160645,
+      "step": 363
+    },
+    {
+      "epoch": 0.11064670567672316,
+      "grad_norm": 2.150822877883911,
+      "learning_rate": 0.00015,
+      "loss": 4.8066911697387695,
+      "step": 364
+    },
+    {
+      "epoch": 0.110950680142868,
+      "grad_norm": 2.0651166439056396,
+      "learning_rate": 0.00015,
+      "loss": 4.713856220245361,
+      "step": 365
+    },
+    {
+      "epoch": 0.11125465460901285,
+      "grad_norm": 2.0583834648132324,
+      "learning_rate": 0.00015,
+      "loss": 4.959167003631592,
+      "step": 366
+    },
+    {
+      "epoch": 0.11155862907515769,
+      "grad_norm": 2.615971803665161,
+      "learning_rate": 0.00015,
+      "loss": 5.0680389404296875,
+      "step": 367
+    },
+    {
+      "epoch": 0.11186260354130254,
+      "grad_norm": 2.209761142730713,
+      "learning_rate": 0.00015,
+      "loss": 4.945580005645752,
+      "step": 368
+    },
+    {
+      "epoch": 0.11216657800744738,
+      "grad_norm": 3.0851430892944336,
+      "learning_rate": 0.00015,
+      "loss": 4.914002418518066,
+      "step": 369
+    },
+    {
+      "epoch": 0.11247055247359222,
+      "grad_norm": 2.8634145259857178,
+      "learning_rate": 0.00015,
+      "loss": 4.797061920166016,
+      "step": 370
+    },
+    {
+      "epoch": 0.11277452693973707,
+      "grad_norm": 2.6660618782043457,
+      "learning_rate": 0.00015,
+      "loss": 4.733559608459473,
+      "step": 371
+    },
+    {
+      "epoch": 0.11307850140588191,
+      "grad_norm": 2.6687684059143066,
+      "learning_rate": 0.00015,
+      "loss": 5.138906002044678,
+      "step": 372
+    },
+    {
+      "epoch": 0.11338247587202675,
+      "grad_norm": 4.277584552764893,
+      "learning_rate": 0.00015,
+      "loss": 5.0002827644348145,
+      "step": 373
+    },
+    {
+      "epoch": 0.1136864503381716,
+      "grad_norm": 3.003919839859009,
+      "learning_rate": 0.00015,
+      "loss": 4.83658504486084,
+      "step": 374
+    },
+    {
+      "epoch": 0.11399042480431644,
+      "grad_norm": 3.2129383087158203,
+      "learning_rate": 0.00015,
+      "loss": 4.9096808433532715,
+      "step": 375
+    },
+    {
+      "epoch": 0.11429439927046128,
+      "grad_norm": 2.587614059448242,
+      "learning_rate": 0.00015,
+      "loss": 4.969578742980957,
+      "step": 376
+    },
+    {
+      "epoch": 0.11459837373660613,
+      "grad_norm": 3.215866804122925,
+      "learning_rate": 0.00015,
+      "loss": 4.7802348136901855,
+      "step": 377
+    },
+    {
+      "epoch": 0.11490234820275097,
+      "grad_norm": 2.8977437019348145,
+      "learning_rate": 0.00015,
+      "loss": 4.842691421508789,
+      "step": 378
+    },
+    {
+      "epoch": 0.11520632266889581,
+      "grad_norm": 1.8513367176055908,
+      "learning_rate": 0.00015,
+      "loss": 4.949589729309082,
+      "step": 379
+    },
+    {
+      "epoch": 0.11551029713504066,
+      "grad_norm": 3.150493860244751,
+      "learning_rate": 0.00015,
+      "loss": 5.150163650512695,
+      "step": 380
+    },
+    {
+      "epoch": 0.1158142716011855,
+      "grad_norm": 1.9769551753997803,
+      "learning_rate": 0.00015,
+      "loss": 5.301183223724365,
+      "step": 381
+    },
+    {
+      "epoch": 0.11611824606733034,
+      "grad_norm": 4.124518394470215,
+      "learning_rate": 0.00015,
+      "loss": 5.31982421875,
+      "step": 382
+    },
+    {
+      "epoch": 0.11642222053347519,
+      "grad_norm": 3.635239601135254,
+      "learning_rate": 0.00015,
+      "loss": 4.8040008544921875,
+      "step": 383
+    },
+    {
+      "epoch": 0.11672619499962003,
+      "grad_norm": 2.596153974533081,
+      "learning_rate": 0.00015,
+      "loss": 4.824268341064453,
+      "step": 384
+    },
+    {
+      "epoch": 0.11703016946576487,
+      "grad_norm": 2.3842713832855225,
+      "learning_rate": 0.00015,
+      "loss": 4.73667573928833,
+      "step": 385
+    },
+    {
+      "epoch": 0.11733414393190972,
+      "grad_norm": 3.162278175354004,
+      "learning_rate": 0.00015,
+      "loss": 4.850276947021484,
+      "step": 386
+    },
+    {
+      "epoch": 0.11763811839805456,
+      "grad_norm": 3.4372119903564453,
+      "learning_rate": 0.00015,
+      "loss": 5.129795074462891,
+      "step": 387
+    },
+    {
+      "epoch": 0.1179420928641994,
+      "grad_norm": 2.546644926071167,
+      "learning_rate": 0.00015,
+      "loss": 4.921144485473633,
+      "step": 388
+    },
+    {
+      "epoch": 0.11824606733034425,
+      "grad_norm": 2.473304033279419,
+      "learning_rate": 0.00015,
+      "loss": 4.992341041564941,
+      "step": 389
+    },
+    {
+      "epoch": 0.11855004179648909,
+      "grad_norm": 2.2003347873687744,
+      "learning_rate": 0.00015,
+      "loss": 4.9349188804626465,
+      "step": 390
+    },
+    {
+      "epoch": 0.11885401626263394,
+      "grad_norm": 2.101527214050293,
+      "learning_rate": 0.00015,
+      "loss": 4.893990993499756,
+      "step": 391
+    },
+    {
+      "epoch": 0.11915799072877878,
+      "grad_norm": 2.4259021282196045,
+      "learning_rate": 0.00015,
+      "loss": 4.635498046875,
+      "step": 392
+    },
+    {
+      "epoch": 0.11946196519492362,
+      "grad_norm": 2.4046413898468018,
+      "learning_rate": 0.00015,
+      "loss": 4.8616943359375,
+      "step": 393
+    },
+    {
+      "epoch": 0.11976593966106847,
+      "grad_norm": 3.7164700031280518,
+      "learning_rate": 0.00015,
+      "loss": 4.59390926361084,
+      "step": 394
+    },
+    {
+      "epoch": 0.12006991412721331,
+      "grad_norm": 2.802246332168579,
+      "learning_rate": 0.00015,
+      "loss": 4.769420146942139,
+      "step": 395
+    },
+    {
+      "epoch": 0.12037388859335815,
+      "grad_norm": 2.2997653484344482,
+      "learning_rate": 0.00015,
+      "loss": 4.942673683166504,
+      "step": 396
+    },
+    {
+      "epoch": 0.120677863059503,
+      "grad_norm": 2.409381866455078,
+      "learning_rate": 0.00015,
+      "loss": 4.540895462036133,
+      "step": 397
+    },
+    {
+      "epoch": 0.12098183752564784,
+      "grad_norm": 2.185817003250122,
+      "learning_rate": 0.00015,
+      "loss": 5.199550628662109,
+      "step": 398
+    },
+    {
+      "epoch": 0.12128581199179268,
+      "grad_norm": 2.232225179672241,
+      "learning_rate": 0.00015,
+      "loss": 4.576483726501465,
+      "step": 399
+    },
+    {
+      "epoch": 0.12158978645793753,
+      "grad_norm": 2.292133331298828,
+      "learning_rate": 0.00015,
+      "loss": 4.864506721496582,
+      "step": 400
+    },
+    {
+      "epoch": 0.12189376092408238,
+      "grad_norm": 2.574544906616211,
+      "learning_rate": 0.00015,
+      "loss": 4.697381973266602,
+      "step": 401
+    },
+    {
+      "epoch": 0.12219773539022723,
+      "grad_norm": 2.691091775894165,
+      "learning_rate": 0.00015,
+      "loss": 4.926554203033447,
+      "step": 402
+    },
+    {
+      "epoch": 0.12250170985637207,
+      "grad_norm": 2.3297572135925293,
+      "learning_rate": 0.00015,
+      "loss": 4.731498718261719,
+      "step": 403
+    },
+    {
+      "epoch": 0.12280568432251691,
+      "grad_norm": 1.9460726976394653,
+      "learning_rate": 0.00015,
+      "loss": 4.978002548217773,
+      "step": 404
+    },
+    {
+      "epoch": 0.12310965878866176,
+      "grad_norm": 2.838120937347412,
+      "learning_rate": 0.00015,
+      "loss": 4.538815498352051,
+      "step": 405
+    },
+    {
+      "epoch": 0.1234136332548066,
+      "grad_norm": 2.720762252807617,
+      "learning_rate": 0.00015,
+      "loss": 5.014606952667236,
+      "step": 406
+    },
+    {
+      "epoch": 0.12371760772095144,
+      "grad_norm": 2.1209354400634766,
+      "learning_rate": 0.00015,
+      "loss": 4.944110870361328,
+      "step": 407
+    },
+    {
+      "epoch": 0.12402158218709629,
+      "grad_norm": 1.9998852014541626,
+      "learning_rate": 0.00015,
+      "loss": 4.702911376953125,
+      "step": 408
+    },
+    {
+      "epoch": 0.12432555665324113,
+      "grad_norm": 2.3556227684020996,
+      "learning_rate": 0.00015,
+      "loss": 4.830615043640137,
+      "step": 409
+    },
+    {
+      "epoch": 0.12462953111938598,
+      "grad_norm": 2.110801935195923,
+      "learning_rate": 0.00015,
+      "loss": 4.796296119689941,
+      "step": 410
+    },
+    {
+      "epoch": 0.12493350558553082,
+      "grad_norm": 2.3564796447753906,
+      "learning_rate": 0.00015,
+      "loss": 4.573785781860352,
+      "step": 411
+    },
+    {
+      "epoch": 0.12523748005167565,
+      "grad_norm": 2.4335200786590576,
+      "learning_rate": 0.00015,
+      "loss": 4.79954719543457,
+      "step": 412
+    },
+    {
+      "epoch": 0.1255414545178205,
+      "grad_norm": 2.158512830734253,
+      "learning_rate": 0.00015,
+      "loss": 4.756845474243164,
+      "step": 413
+    },
+    {
+      "epoch": 0.12584542898396534,
+      "grad_norm": 2.4842419624328613,
+      "learning_rate": 0.00015,
+      "loss": 4.982643127441406,
+      "step": 414
+    },
+    {
+      "epoch": 0.1261494034501102,
+      "grad_norm": 2.4078800678253174,
+      "learning_rate": 0.00015,
+      "loss": 4.6305975914001465,
+      "step": 415
+    },
+    {
+      "epoch": 0.12645337791625502,
+      "grad_norm": 2.561401605606079,
+      "learning_rate": 0.00015,
+      "loss": 5.108941555023193,
+      "step": 416
+    },
+    {
+      "epoch": 0.12675735238239988,
+      "grad_norm": 1.8021341562271118,
+      "learning_rate": 0.00015,
+      "loss": 4.702005863189697,
+      "step": 417
+    },
+    {
+      "epoch": 0.1270613268485447,
+      "grad_norm": 2.0056488513946533,
+      "learning_rate": 0.00015,
+      "loss": 4.661421775817871,
+      "step": 418
+    },
+    {
+      "epoch": 0.12736530131468957,
+      "grad_norm": 3.382174015045166,
+      "learning_rate": 0.00015,
+      "loss": 5.086690902709961,
+      "step": 419
+    },
+    {
+      "epoch": 0.12766927578083442,
+      "grad_norm": 2.234229803085327,
+      "learning_rate": 0.00015,
+      "loss": 4.989408493041992,
+      "step": 420
+    },
+    {
+      "epoch": 0.12797325024697925,
+      "grad_norm": 2.2572667598724365,
+      "learning_rate": 0.00015,
+      "loss": 4.917817115783691,
+      "step": 421
+    },
+    {
+      "epoch": 0.1282772247131241,
+      "grad_norm": 2.1507554054260254,
+      "learning_rate": 0.00015,
+      "loss": 5.058313369750977,
+      "step": 422
+    },
+    {
+      "epoch": 0.12858119917926894,
+      "grad_norm": 2.303370237350464,
+      "learning_rate": 0.00015,
+      "loss": 4.646108627319336,
+      "step": 423
+    },
+    {
+      "epoch": 0.1288851736454138,
+      "grad_norm": 2.0629470348358154,
+      "learning_rate": 0.00015,
+      "loss": 4.770839691162109,
+      "step": 424
+    },
+    {
+      "epoch": 0.12918914811155863,
+      "grad_norm": 2.3262033462524414,
+      "learning_rate": 0.00015,
+      "loss": 5.0411882400512695,
+      "step": 425
+    },
+    {
+      "epoch": 0.12949312257770348,
+      "grad_norm": 2.294067144393921,
+      "learning_rate": 0.00015,
+      "loss": 4.742310047149658,
+      "step": 426
+    },
+    {
+      "epoch": 0.12979709704384831,
+      "grad_norm": 2.285543441772461,
+      "learning_rate": 0.00015,
+      "loss": 5.011312484741211,
+      "step": 427
+    },
+    {
+      "epoch": 0.13010107150999317,
+      "grad_norm": 2.660085439682007,
+      "learning_rate": 0.00015,
+      "loss": 5.061105728149414,
+      "step": 428
+    },
+    {
+      "epoch": 0.130405045976138,
+      "grad_norm": 1.7300945520401,
+      "learning_rate": 0.00015,
+      "loss": 4.713881492614746,
+      "step": 429
+    },
+    {
+      "epoch": 0.13070902044228286,
+      "grad_norm": 2.3755970001220703,
+      "learning_rate": 0.00015,
+      "loss": 4.735212326049805,
+      "step": 430
+    },
+    {
+      "epoch": 0.1310129949084277,
+      "grad_norm": 2.724358558654785,
+      "learning_rate": 0.00015,
+      "loss": 4.866620063781738,
+      "step": 431
+    },
+    {
+      "epoch": 0.13131696937457255,
+      "grad_norm": 1.9843202829360962,
+      "learning_rate": 0.00015,
+      "loss": 4.900468826293945,
+      "step": 432
+    },
+    {
+      "epoch": 0.13162094384071737,
+      "grad_norm": 2.1709346771240234,
+      "learning_rate": 0.00015,
+      "loss": 4.884383201599121,
+      "step": 433
+    },
+    {
+      "epoch": 0.13192491830686223,
+      "grad_norm": 1.9553192853927612,
+      "learning_rate": 0.00015,
+      "loss": 4.677547931671143,
+      "step": 434
+    },
+    {
+      "epoch": 0.13222889277300706,
+      "grad_norm": 2.137723684310913,
+      "learning_rate": 0.00015,
+      "loss": 4.841003894805908,
+      "step": 435
+    },
+    {
+      "epoch": 0.13253286723915192,
+      "grad_norm": 2.5220134258270264,
+      "learning_rate": 0.00015,
+      "loss": 4.728403091430664,
+      "step": 436
+    },
+    {
+      "epoch": 0.13283684170529675,
+      "grad_norm": 2.3277413845062256,
+      "learning_rate": 0.00015,
+      "loss": 4.5765380859375,
+      "step": 437
+    },
+    {
+      "epoch": 0.1331408161714416,
+      "grad_norm": 1.8570560216903687,
+      "learning_rate": 0.00015,
+      "loss": 4.832406997680664,
+      "step": 438
+    },
+    {
+      "epoch": 0.13344479063758644,
+      "grad_norm": 2.1348214149475098,
+      "learning_rate": 0.00015,
+      "loss": 4.744243621826172,
+      "step": 439
+    },
+    {
+      "epoch": 0.1337487651037313,
+      "grad_norm": 2.06382417678833,
+      "learning_rate": 0.00015,
+      "loss": 4.859317779541016,
+      "step": 440
+    },
+    {
+      "epoch": 0.13405273956987612,
+      "grad_norm": 2.34126615524292,
+      "learning_rate": 0.00015,
+      "loss": 4.874665260314941,
+      "step": 441
+    },
+    {
+      "epoch": 0.13435671403602098,
+      "grad_norm": 2.6390693187713623,
+      "learning_rate": 0.00015,
+      "loss": 5.0122880935668945,
+      "step": 442
+    },
+    {
+      "epoch": 0.1346606885021658,
+      "grad_norm": 1.9252774715423584,
+      "learning_rate": 0.00015,
+      "loss": 4.924732685089111,
+      "step": 443
+    },
+    {
+      "epoch": 0.13496466296831067,
+      "grad_norm": 2.233944892883301,
+      "learning_rate": 0.00015,
+      "loss": 5.066226482391357,
+      "step": 444
+    },
+    {
+      "epoch": 0.1352686374344555,
+      "grad_norm": 2.1682188510894775,
+      "learning_rate": 0.00015,
+      "loss": 4.715705394744873,
+      "step": 445
+    },
+    {
+      "epoch": 0.13557261190060035,
+      "grad_norm": 2.2908785343170166,
+      "learning_rate": 0.00015,
+      "loss": 5.011838436126709,
+      "step": 446
+    },
+    {
+      "epoch": 0.13587658636674518,
+      "grad_norm": 2.2433042526245117,
+      "learning_rate": 0.00015,
+      "loss": 4.785195350646973,
+      "step": 447
+    },
+    {
+      "epoch": 0.13618056083289004,
+      "grad_norm": 1.8833383321762085,
+      "learning_rate": 0.00015,
+      "loss": 4.529738426208496,
+      "step": 448
+    },
+    {
+      "epoch": 0.13648453529903487,
+      "grad_norm": 2.0440526008605957,
+      "learning_rate": 0.00015,
+      "loss": 4.9055891036987305,
+      "step": 449
+    },
+    {
+      "epoch": 0.13678850976517973,
+      "grad_norm": 1.798686146736145,
+      "learning_rate": 0.00015,
+      "loss": 4.654270172119141,
+      "step": 450
+    },
+    {
+      "epoch": 0.13709248423132456,
+      "grad_norm": 1.9728049039840698,
+      "learning_rate": 0.00015,
+      "loss": 4.722278118133545,
+      "step": 451
+    },
+    {
+      "epoch": 0.13739645869746941,
+      "grad_norm": 3.0724036693573,
+      "learning_rate": 0.00015,
+      "loss": 5.004765510559082,
+      "step": 452
+    },
+    {
+      "epoch": 0.13770043316361424,
+      "grad_norm": 2.3341281414031982,
+      "learning_rate": 0.00015,
+      "loss": 4.829780101776123,
+      "step": 453
+    },
+    {
+      "epoch": 0.1380044076297591,
+      "grad_norm": 1.8180235624313354,
+      "learning_rate": 0.00015,
+      "loss": 4.661438941955566,
+      "step": 454
+    },
+    {
+      "epoch": 0.13830838209590393,
+      "grad_norm": 2.3792552947998047,
+      "learning_rate": 0.00015,
+      "loss": 4.527836799621582,
+      "step": 455
+    },
+    {
+      "epoch": 0.1386123565620488,
+      "grad_norm": 2.1522226333618164,
+      "learning_rate": 0.00015,
+      "loss": 4.738117218017578,
+      "step": 456
+    },
+    {
+      "epoch": 0.13891633102819362,
+      "grad_norm": 2.0583791732788086,
+      "learning_rate": 0.00015,
+      "loss": 4.95326042175293,
+      "step": 457
+    },
+    {
+      "epoch": 0.13922030549433848,
+      "grad_norm": 1.9225889444351196,
+      "learning_rate": 0.00015,
+      "loss": 4.537798881530762,
+      "step": 458
+    },
+    {
+      "epoch": 0.13952427996048333,
+      "grad_norm": 1.958419919013977,
+      "learning_rate": 0.00015,
+      "loss": 4.922906875610352,
+      "step": 459
+    },
+    {
+      "epoch": 0.13982825442662816,
+      "grad_norm": 2.788328170776367,
+      "learning_rate": 0.00015,
+      "loss": 4.768488883972168,
+      "step": 460
+    },
+    {
+      "epoch": 0.14013222889277302,
+      "grad_norm": 1.8830108642578125,
+      "learning_rate": 0.00015,
+      "loss": 4.650717735290527,
+      "step": 461
+    },
+    {
+      "epoch": 0.14043620335891785,
+      "grad_norm": 2.232781171798706,
+      "learning_rate": 0.00015,
+      "loss": 4.682582855224609,
+      "step": 462
+    },
+    {
+      "epoch": 0.1407401778250627,
+      "grad_norm": 2.376508951187134,
+      "learning_rate": 0.00015,
+      "loss": 4.626139163970947,
+      "step": 463
+    },
+    {
+      "epoch": 0.14104415229120754,
+      "grad_norm": 2.0443906784057617,
+      "learning_rate": 0.00015,
+      "loss": 4.70169734954834,
+      "step": 464
+    },
+    {
+      "epoch": 0.1413481267573524,
+      "grad_norm": 2.342967987060547,
+      "learning_rate": 0.00015,
+      "loss": 4.964600086212158,
+      "step": 465
+    },
+    {
+      "epoch": 0.14165210122349722,
+      "grad_norm": 2.307633876800537,
+      "learning_rate": 0.00015,
+      "loss": 4.552054405212402,
+      "step": 466
+    },
+    {
+      "epoch": 0.14195607568964208,
+      "grad_norm": 2.3887476921081543,
+      "learning_rate": 0.00015,
+      "loss": 4.6841278076171875,
+      "step": 467
+    },
+    {
+      "epoch": 0.1422600501557869,
+      "grad_norm": 2.263071298599243,
+      "learning_rate": 0.00015,
+      "loss": 4.890757083892822,
+      "step": 468
+    },
+    {
+      "epoch": 0.14256402462193177,
+      "grad_norm": 2.154442071914673,
+      "learning_rate": 0.00015,
+      "loss": 4.540566444396973,
+      "step": 469
+    },
+    {
+      "epoch": 0.1428679990880766,
+      "grad_norm": 2.0220494270324707,
+      "learning_rate": 0.00015,
+      "loss": 4.652443885803223,
+      "step": 470
+    },
+    {
+      "epoch": 0.14317197355422145,
+      "grad_norm": 2.3018667697906494,
+      "learning_rate": 0.00015,
+      "loss": 4.67112922668457,
+      "step": 471
+    },
+    {
+      "epoch": 0.14347594802036628,
+      "grad_norm": 3.268699884414673,
+      "learning_rate": 0.00015,
+      "loss": 5.474539756774902,
+      "step": 472
+    },
+    {
+      "epoch": 0.14377992248651114,
+      "grad_norm": 2.5835506916046143,
+      "learning_rate": 0.00015,
+      "loss": 4.545164585113525,
+      "step": 473
+    },
+    {
+      "epoch": 0.14408389695265597,
+      "grad_norm": 2.4063456058502197,
+      "learning_rate": 0.00015,
+      "loss": 5.027111053466797,
+      "step": 474
+    },
+    {
+      "epoch": 0.14438787141880083,
+      "grad_norm": 1.9622764587402344,
+      "learning_rate": 0.00015,
+      "loss": 4.427340030670166,
+      "step": 475
+    },
+    {
+      "epoch": 0.14469184588494566,
+      "grad_norm": 2.7230958938598633,
+      "learning_rate": 0.00015,
+      "loss": 4.779913902282715,
+      "step": 476
+    },
+    {
+      "epoch": 0.14499582035109052,
+      "grad_norm": 2.4062132835388184,
+      "learning_rate": 0.00015,
+      "loss": 4.742057800292969,
+      "step": 477
+    },
+    {
+      "epoch": 0.14529979481723534,
+      "grad_norm": 2.3851118087768555,
+      "learning_rate": 0.00015,
+      "loss": 4.711178302764893,
+      "step": 478
+    },
+    {
+      "epoch": 0.1456037692833802,
+      "grad_norm": 2.240133047103882,
+      "learning_rate": 0.00015,
+      "loss": 4.920535564422607,
+      "step": 479
+    },
+    {
+      "epoch": 0.14590774374952503,
+      "grad_norm": 2.181083917617798,
+      "learning_rate": 0.00015,
+      "loss": 4.7905473709106445,
+      "step": 480
+    },
+    {
+      "epoch": 0.1462117182156699,
+      "grad_norm": 2.741213321685791,
+      "learning_rate": 0.00015,
+      "loss": 4.928895950317383,
+      "step": 481
+    },
+    {
+      "epoch": 0.14651569268181472,
+      "grad_norm": 2.198018789291382,
+      "learning_rate": 0.00015,
+      "loss": 4.798812389373779,
+      "step": 482
+    },
+    {
+      "epoch": 0.14681966714795958,
+      "grad_norm": 1.908618688583374,
+      "learning_rate": 0.00015,
+      "loss": 4.615466117858887,
+      "step": 483
+    },
+    {
+      "epoch": 0.1471236416141044,
+      "grad_norm": 2.6483750343322754,
+      "learning_rate": 0.00015,
+      "loss": 4.656067371368408,
+      "step": 484
+    },
+    {
+      "epoch": 0.14742761608024926,
+      "grad_norm": 2.4816155433654785,
+      "learning_rate": 0.00015,
+      "loss": 5.035987854003906,
+      "step": 485
+    },
+    {
+      "epoch": 0.1477315905463941,
+      "grad_norm": 1.8377639055252075,
+      "learning_rate": 0.00015,
+      "loss": 4.701920509338379,
+      "step": 486
+    },
+    {
+      "epoch": 0.14803556501253895,
+      "grad_norm": 2.5602715015411377,
+      "learning_rate": 0.00015,
+      "loss": 5.001008987426758,
+      "step": 487
+    },
+    {
+      "epoch": 0.14833953947868378,
+      "grad_norm": 1.9801889657974243,
+      "learning_rate": 0.00015,
+      "loss": 4.687620162963867,
+      "step": 488
+    },
+    {
+      "epoch": 0.14864351394482864,
+      "grad_norm": 3.2698171138763428,
+      "learning_rate": 0.00015,
+      "loss": 5.092370986938477,
+      "step": 489
+    },
+    {
+      "epoch": 0.14894748841097347,
+      "grad_norm": 1.9460852146148682,
+      "learning_rate": 0.00015,
+      "loss": 4.615166664123535,
+      "step": 490
+    },
+    {
+      "epoch": 0.14925146287711832,
+      "grad_norm": 1.9579049348831177,
+      "learning_rate": 0.00015,
+      "loss": 4.757729530334473,
+      "step": 491
+    },
+    {
+      "epoch": 0.14955543734326315,
+      "grad_norm": 2.406999349594116,
+      "learning_rate": 0.00015,
+      "loss": 4.645845413208008,
+      "step": 492
+    },
+    {
+      "epoch": 0.149859411809408,
+      "grad_norm": 2.391007423400879,
+      "learning_rate": 0.00015,
+      "loss": 5.1390228271484375,
+      "step": 493
+    },
+    {
+      "epoch": 0.15016338627555284,
+      "grad_norm": 2.904451608657837,
+      "learning_rate": 0.00015,
+      "loss": 4.755255699157715,
+      "step": 494
+    },
+    {
+      "epoch": 0.1504673607416977,
+      "grad_norm": 2.1067545413970947,
+      "learning_rate": 0.00015,
+      "loss": 4.736544609069824,
+      "step": 495
+    },
+    {
+      "epoch": 0.15077133520784256,
+      "grad_norm": 2.3628900051116943,
+      "learning_rate": 0.00015,
+      "loss": 5.111974716186523,
+      "step": 496
+    },
+    {
+      "epoch": 0.15107530967398738,
+      "grad_norm": 2.7027862071990967,
+      "learning_rate": 0.00015,
+      "loss": 4.868869304656982,
+      "step": 497
+    },
+    {
+      "epoch": 0.15137928414013224,
+      "grad_norm": 2.2270071506500244,
+      "learning_rate": 0.00015,
+      "loss": 4.804258346557617,
+      "step": 498
+    },
+    {
+      "epoch": 0.15168325860627707,
+      "grad_norm": 2.1076321601867676,
+      "learning_rate": 0.00015,
+      "loss": 4.735129356384277,
+      "step": 499
+    },
+    {
+      "epoch": 0.15198723307242193,
+      "grad_norm": 2.232858419418335,
+      "learning_rate": 0.00015,
+      "loss": 4.720833778381348,
+      "step": 500
+    },
+    {
+      "epoch": 0.15229120753856676,
+      "grad_norm": 1.8920719623565674,
+      "learning_rate": 0.00015,
+      "loss": 4.741750717163086,
+      "step": 501
+    },
+    {
+      "epoch": 0.15259518200471162,
+      "grad_norm": 2.012943983078003,
+      "learning_rate": 0.00015,
+      "loss": 4.64883279800415,
+      "step": 502
+    },
+    {
+      "epoch": 0.15289915647085645,
+      "grad_norm": 2.114447593688965,
+      "learning_rate": 0.00015,
+      "loss": 4.6952314376831055,
+      "step": 503
+    },
+    {
+      "epoch": 0.1532031309370013,
+      "grad_norm": 2.0654213428497314,
+      "learning_rate": 0.00015,
+      "loss": 4.549853324890137,
+      "step": 504
+    },
+    {
+      "epoch": 0.15350710540314613,
+      "grad_norm": 2.710348606109619,
+      "learning_rate": 0.00015,
+      "loss": 4.760828971862793,
+      "step": 505
+    },
+    {
+      "epoch": 0.153811079869291,
+      "grad_norm": 2.310258626937866,
+      "learning_rate": 0.00015,
+      "loss": 4.656004905700684,
+      "step": 506
+    },
+    {
+      "epoch": 0.15411505433543582,
+      "grad_norm": 1.7558186054229736,
+      "learning_rate": 0.00015,
+      "loss": 4.6995649337768555,
+      "step": 507
+    },
+    {
+      "epoch": 0.15441902880158068,
+      "grad_norm": 2.1797780990600586,
+      "learning_rate": 0.00015,
+      "loss": 4.522988319396973,
+      "step": 508
+    },
+    {
+      "epoch": 0.1547230032677255,
+      "grad_norm": 2.1816389560699463,
+      "learning_rate": 0.00015,
+      "loss": 4.843192100524902,
+      "step": 509
+    },
+    {
+      "epoch": 0.15502697773387036,
+      "grad_norm": 2.8029768466949463,
+      "learning_rate": 0.00015,
+      "loss": 4.8778581619262695,
+      "step": 510
+    },
+    {
+      "epoch": 0.1553309522000152,
+      "grad_norm": 2.006779670715332,
+      "learning_rate": 0.00015,
+      "loss": 4.529073715209961,
+      "step": 511
+    },
+    {
+      "epoch": 0.15563492666616005,
+      "grad_norm": 2.316190004348755,
+      "learning_rate": 0.00015,
+      "loss": 4.966833114624023,
+      "step": 512
+    },
+    {
+      "epoch": 0.15593890113230488,
+      "grad_norm": 2.7688729763031006,
+      "learning_rate": 0.00015,
+      "loss": 4.617415904998779,
+      "step": 513
+    },
+    {
+      "epoch": 0.15624287559844974,
+      "grad_norm": 1.8253449201583862,
+      "learning_rate": 0.00015,
+      "loss": 4.509947776794434,
+      "step": 514
+    },
+    {
+      "epoch": 0.15654685006459457,
+      "grad_norm": 2.077517032623291,
+      "learning_rate": 0.00015,
+      "loss": 4.806764602661133,
+      "step": 515
+    },
+    {
+      "epoch": 0.15685082453073942,
+      "grad_norm": 2.3121280670166016,
+      "learning_rate": 0.00015,
+      "loss": 4.502899646759033,
+      "step": 516
+    },
+    {
+      "epoch": 0.15715479899688425,
+      "grad_norm": 2.1371910572052,
+      "learning_rate": 0.00015,
+      "loss": 4.636500358581543,
+      "step": 517
+    },
+    {
+      "epoch": 0.1574587734630291,
+      "grad_norm": 2.4253716468811035,
+      "learning_rate": 0.00015,
+      "loss": 4.4380927085876465,
+      "step": 518
+    },
+    {
+      "epoch": 0.15776274792917394,
+      "grad_norm": 1.9075349569320679,
+      "learning_rate": 0.00015,
+      "loss": 4.7069902420043945,
+      "step": 519
+    },
+    {
+      "epoch": 0.1580667223953188,
+      "grad_norm": 2.134714126586914,
+      "learning_rate": 0.00015,
+      "loss": 4.796967506408691,
+      "step": 520
+    },
+    {
+      "epoch": 0.15837069686146363,
+      "grad_norm": 2.418729543685913,
+      "learning_rate": 0.00015,
+      "loss": 4.645528793334961,
+      "step": 521
+    },
+    {
+      "epoch": 0.15867467132760849,
+      "grad_norm": 2.0582802295684814,
+      "learning_rate": 0.00015,
+      "loss": 4.797185897827148,
+      "step": 522
+    },
+    {
+      "epoch": 0.15897864579375331,
+      "grad_norm": 2.4252965450286865,
+      "learning_rate": 0.00015,
+      "loss": 5.015091896057129,
+      "step": 523
+    },
+    {
+      "epoch": 0.15928262025989817,
+      "grad_norm": 2.275078296661377,
+      "learning_rate": 0.00015,
+      "loss": 4.651376724243164,
+      "step": 524
+    },
+    {
+      "epoch": 0.159586594726043,
+      "grad_norm": 2.458658456802368,
+      "learning_rate": 0.00015,
+      "loss": 4.951380729675293,
+      "step": 525
+    },
+    {
+      "epoch": 0.15989056919218786,
+      "grad_norm": 1.959053635597229,
+      "learning_rate": 0.00015,
+      "loss": 4.594791412353516,
+      "step": 526
+    },
+    {
+      "epoch": 0.1601945436583327,
+      "grad_norm": 2.03511381149292,
+      "learning_rate": 0.00015,
+      "loss": 5.107573509216309,
+      "step": 527
+    },
+    {
+      "epoch": 0.16049851812447755,
+      "grad_norm": 1.888108491897583,
+      "learning_rate": 0.00015,
+      "loss": 4.699661731719971,
+      "step": 528
+    },
+    {
+      "epoch": 0.16080249259062238,
+      "grad_norm": 2.417900562286377,
+      "learning_rate": 0.00015,
+      "loss": 4.900885581970215,
+      "step": 529
+    },
+    {
+      "epoch": 0.16110646705676723,
+      "grad_norm": 1.9693663120269775,
+      "learning_rate": 0.00015,
+      "loss": 4.845917701721191,
+      "step": 530
+    },
+    {
+      "epoch": 0.16141044152291206,
+      "grad_norm": 2.4869401454925537,
+      "learning_rate": 0.00015,
+      "loss": 5.013619422912598,
+      "step": 531
+    },
+    {
+      "epoch": 0.16171441598905692,
+      "grad_norm": 2.9126336574554443,
+      "learning_rate": 0.00015,
+      "loss": 4.889483451843262,
+      "step": 532
+    },
+    {
+      "epoch": 0.16201839045520175,
+      "grad_norm": 2.6625611782073975,
+      "learning_rate": 0.00015,
+      "loss": 4.625283241271973,
+      "step": 533
+    },
+    {
+      "epoch": 0.1623223649213466,
+      "grad_norm": 2.2297239303588867,
+      "learning_rate": 0.00015,
+      "loss": 4.867910385131836,
+      "step": 534
+    },
+    {
+      "epoch": 0.16262633938749146,
+      "grad_norm": 2.3175439834594727,
+      "learning_rate": 0.00015,
+      "loss": 4.826192855834961,
+      "step": 535
+    },
+    {
+      "epoch": 0.1629303138536363,
+      "grad_norm": 2.117565631866455,
+      "learning_rate": 0.00015,
+      "loss": 4.662595272064209,
+      "step": 536
+    },
+    {
+      "epoch": 0.16323428831978115,
+      "grad_norm": 2.1798038482666016,
+      "learning_rate": 0.00015,
+      "loss": 4.657549858093262,
+      "step": 537
+    },
+    {
+      "epoch": 0.16353826278592598,
+      "grad_norm": 1.940258264541626,
+      "learning_rate": 0.00015,
+      "loss": 4.978257179260254,
+      "step": 538
+    },
+    {
+      "epoch": 0.16384223725207084,
+      "grad_norm": 2.6843626499176025,
+      "learning_rate": 0.00015,
+      "loss": 4.491443634033203,
+      "step": 539
+    },
+    {
+      "epoch": 0.16414621171821567,
+      "grad_norm": 2.5011868476867676,
+      "learning_rate": 0.00015,
+      "loss": 4.638598442077637,
+      "step": 540
+    },
+    {
+      "epoch": 0.16445018618436053,
+      "grad_norm": 1.8307850360870361,
+      "learning_rate": 0.00015,
+      "loss": 4.797595977783203,
+      "step": 541
+    },
+    {
+      "epoch": 0.16475416065050535,
+      "grad_norm": 2.2151477336883545,
+      "learning_rate": 0.00015,
+      "loss": 4.751269817352295,
+      "step": 542
+    },
+    {
+      "epoch": 0.1650581351166502,
+      "grad_norm": 2.2703468799591064,
+      "learning_rate": 0.00015,
+      "loss": 4.968122482299805,
+      "step": 543
+    },
+    {
+      "epoch": 0.16536210958279504,
+      "grad_norm": 2.0396437644958496,
+      "learning_rate": 0.00015,
+      "loss": 4.7669854164123535,
+      "step": 544
+    },
+    {
+      "epoch": 0.1656660840489399,
+      "grad_norm": 2.384148359298706,
+      "learning_rate": 0.00015,
+      "loss": 4.46330451965332,
+      "step": 545
+    },
+    {
+      "epoch": 0.16597005851508473,
+      "grad_norm": 2.3803882598876953,
+      "learning_rate": 0.00015,
+      "loss": 4.761402130126953,
+      "step": 546
+    },
+    {
+      "epoch": 0.16627403298122959,
+      "grad_norm": 2.031144380569458,
+      "learning_rate": 0.00015,
+      "loss": 4.651483058929443,
+      "step": 547
+    },
+    {
+      "epoch": 0.16657800744737442,
+      "grad_norm": 2.1903669834136963,
+      "learning_rate": 0.00015,
+      "loss": 4.774244785308838,
+      "step": 548
+    },
+    {
+      "epoch": 0.16688198191351927,
+      "grad_norm": 1.934736967086792,
+      "learning_rate": 0.00015,
+      "loss": 4.771899223327637,
+      "step": 549
+    },
+    {
+      "epoch": 0.1671859563796641,
+      "grad_norm": 2.770162582397461,
+      "learning_rate": 0.00015,
+      "loss": 4.729818820953369,
+      "step": 550
+    },
+    {
+      "epoch": 0.16748993084580896,
+      "grad_norm": 1.771448016166687,
+      "learning_rate": 0.00015,
+      "loss": 4.699235916137695,
+      "step": 551
+    },
+    {
+      "epoch": 0.1677939053119538,
+      "grad_norm": 1.9773725271224976,
+      "learning_rate": 0.00015,
+      "loss": 4.9152679443359375,
+      "step": 552
+    },
+    {
+      "epoch": 0.16809787977809865,
+      "grad_norm": 2.7509164810180664,
+      "learning_rate": 0.00015,
+      "loss": 4.846560001373291,
+      "step": 553
+    },
+    {
+      "epoch": 0.16840185424424348,
+      "grad_norm": 1.8572801351547241,
+      "learning_rate": 0.00015,
+      "loss": 4.639094829559326,
+      "step": 554
+    },
+    {
+      "epoch": 0.16870582871038833,
+      "grad_norm": 1.996209979057312,
+      "learning_rate": 0.00015,
+      "loss": 4.905031204223633,
+      "step": 555
+    },
+    {
+      "epoch": 0.16900980317653316,
+      "grad_norm": 2.2827439308166504,
+      "learning_rate": 0.00015,
+      "loss": 4.652318477630615,
+      "step": 556
+    },
+    {
+      "epoch": 0.16931377764267802,
+      "grad_norm": 1.790662169456482,
+      "learning_rate": 0.00015,
+      "loss": 4.632669925689697,
+      "step": 557
+    },
+    {
+      "epoch": 0.16961775210882285,
+      "grad_norm": 1.9209315776824951,
+      "learning_rate": 0.00015,
+      "loss": 4.577767848968506,
+      "step": 558
+    },
+    {
+      "epoch": 0.1699217265749677,
+      "grad_norm": 1.5667444467544556,
+      "learning_rate": 0.00015,
+      "loss": 4.702671051025391,
+      "step": 559
+    },
+    {
+      "epoch": 0.17022570104111254,
+      "grad_norm": 2.354146957397461,
+      "learning_rate": 0.00015,
+      "loss": 4.677872180938721,
+      "step": 560
+    },
+    {
+      "epoch": 0.1705296755072574,
+      "grad_norm": 2.4916112422943115,
+      "learning_rate": 0.00015,
+      "loss": 4.665243148803711,
+      "step": 561
+    },
+    {
+      "epoch": 0.17083364997340222,
+      "grad_norm": 2.3338184356689453,
+      "learning_rate": 0.00015,
+      "loss": 4.356175422668457,
+      "step": 562
+    },
+    {
+      "epoch": 0.17113762443954708,
+      "grad_norm": 3.170602321624756,
+      "learning_rate": 0.00015,
+      "loss": 4.678605556488037,
+      "step": 563
+    },
+    {
+      "epoch": 0.1714415989056919,
+      "grad_norm": 2.9348208904266357,
+      "learning_rate": 0.00015,
+      "loss": 4.992365837097168,
+      "step": 564
+    },
+    {
+      "epoch": 0.17174557337183677,
+      "grad_norm": 1.723535180091858,
+      "learning_rate": 0.00015,
+      "loss": 4.653353691101074,
+      "step": 565
+    },
+    {
+      "epoch": 0.1720495478379816,
+      "grad_norm": 3.7177646160125732,
+      "learning_rate": 0.00015,
+      "loss": 4.81765079498291,
+      "step": 566
+    },
+    {
+      "epoch": 0.17235352230412646,
+      "grad_norm": 2.2390098571777344,
+      "learning_rate": 0.00015,
+      "loss": 4.792444229125977,
+      "step": 567
+    },
+    {
+      "epoch": 0.17265749677027128,
+      "grad_norm": 2.2799899578094482,
+      "learning_rate": 0.00015,
+      "loss": 4.390934944152832,
+      "step": 568
+    },
+    {
+      "epoch": 0.17296147123641614,
+      "grad_norm": 2.5163068771362305,
+      "learning_rate": 0.00015,
+      "loss": 4.652558326721191,
+      "step": 569
+    },
+    {
+      "epoch": 0.17326544570256097,
+      "grad_norm": 2.3843750953674316,
+      "learning_rate": 0.00015,
+      "loss": 4.965662956237793,
+      "step": 570
+    },
+    {
+      "epoch": 0.17356942016870583,
+      "grad_norm": 1.8874980211257935,
+      "learning_rate": 0.00015,
+      "loss": 4.386003494262695,
+      "step": 571
+    },
+    {
+      "epoch": 0.17387339463485066,
+      "grad_norm": 2.642117500305176,
+      "learning_rate": 0.00015,
+      "loss": 4.613264083862305,
+      "step": 572
+    },
+    {
+      "epoch": 0.17417736910099552,
+      "grad_norm": 1.8854055404663086,
+      "learning_rate": 0.00015,
+      "loss": 4.5604472160339355,
+      "step": 573
+    },
+    {
+      "epoch": 0.17448134356714037,
+      "grad_norm": 1.7187323570251465,
+      "learning_rate": 0.00015,
+      "loss": 4.515700340270996,
+      "step": 574
+    },
+    {
+      "epoch": 0.1747853180332852,
+      "grad_norm": 1.9935396909713745,
+      "learning_rate": 0.00015,
+      "loss": 4.629323959350586,
+      "step": 575
+    },
+    {
+      "epoch": 0.17508929249943006,
+      "grad_norm": 2.427905321121216,
+      "learning_rate": 0.00015,
+      "loss": 4.779745578765869,
+      "step": 576
+    },
+    {
+      "epoch": 0.1753932669655749,
+      "grad_norm": 2.262659788131714,
+      "learning_rate": 0.00015,
+      "loss": 4.678261756896973,
+      "step": 577
+    },
+    {
+      "epoch": 0.17569724143171975,
+      "grad_norm": 2.2216780185699463,
+      "learning_rate": 0.00015,
+      "loss": 4.771636009216309,
+      "step": 578
+    },
+    {
+      "epoch": 0.17600121589786458,
+      "grad_norm": 2.3558363914489746,
+      "learning_rate": 0.00015,
+      "loss": 4.785818576812744,
+      "step": 579
+    },
+    {
+      "epoch": 0.17630519036400943,
+      "grad_norm": 2.638533115386963,
+      "learning_rate": 0.00015,
+      "loss": 4.6706671714782715,
+      "step": 580
+    },
+    {
+      "epoch": 0.17660916483015426,
+      "grad_norm": 1.9045906066894531,
+      "learning_rate": 0.00015,
+      "loss": 4.697237014770508,
+      "step": 581
+    },
+    {
+      "epoch": 0.17691313929629912,
+      "grad_norm": 2.6674656867980957,
+      "learning_rate": 0.00015,
+      "loss": 4.6974029541015625,
+      "step": 582
+    },
+    {
+      "epoch": 0.17721711376244395,
+      "grad_norm": 2.682739496231079,
+      "learning_rate": 0.00015,
+      "loss": 4.4234418869018555,
+      "step": 583
+    },
+    {
+      "epoch": 0.1775210882285888,
+      "grad_norm": 2.4700725078582764,
+      "learning_rate": 0.00015,
+      "loss": 4.428589820861816,
+      "step": 584
+    },
+    {
+      "epoch": 0.17782506269473364,
+      "grad_norm": 2.50897216796875,
+      "learning_rate": 0.00015,
+      "loss": 4.905554294586182,
+      "step": 585
+    },
+    {
+      "epoch": 0.1781290371608785,
+      "grad_norm": 1.9947826862335205,
+      "learning_rate": 0.00015,
+      "loss": 4.777799606323242,
+      "step": 586
+    },
+    {
+      "epoch": 0.17843301162702332,
+      "grad_norm": 1.8332667350769043,
+      "learning_rate": 0.00015,
+      "loss": 4.642075538635254,
+      "step": 587
+    },
+    {
+      "epoch": 0.17873698609316818,
+      "grad_norm": 2.5155789852142334,
+      "learning_rate": 0.00015,
+      "loss": 4.5264058113098145,
+      "step": 588
+    },
+    {
+      "epoch": 0.179040960559313,
+      "grad_norm": 2.315805673599243,
+      "learning_rate": 0.00015,
+      "loss": 4.999406814575195,
+      "step": 589
+    },
+    {
+      "epoch": 0.17934493502545787,
+      "grad_norm": 1.84408438205719,
+      "learning_rate": 0.00015,
+      "loss": 4.722995758056641,
+      "step": 590
+    },
+    {
+      "epoch": 0.1796489094916027,
+      "grad_norm": 2.8011281490325928,
+      "learning_rate": 0.00015,
+      "loss": 4.465986251831055,
+      "step": 591
+    },
+    {
+      "epoch": 0.17995288395774756,
+      "grad_norm": 1.6783325672149658,
+      "learning_rate": 0.00015,
+      "loss": 4.507539749145508,
+      "step": 592
+    },
+    {
+      "epoch": 0.18025685842389239,
+      "grad_norm": 3.515681743621826,
+      "learning_rate": 0.00015,
+      "loss": 4.765458583831787,
+      "step": 593
+    },
+    {
+      "epoch": 0.18056083289003724,
+      "grad_norm": 2.6943411827087402,
+      "learning_rate": 0.00015,
+      "loss": 4.8780131340026855,
+      "step": 594
+    },
+    {
+      "epoch": 0.18086480735618207,
+      "grad_norm": 2.4187510013580322,
+      "learning_rate": 0.00015,
+      "loss": 4.971892356872559,
+      "step": 595
+    },
+    {
+      "epoch": 0.18116878182232693,
+      "grad_norm": 2.3067448139190674,
+      "learning_rate": 0.00015,
+      "loss": 4.879145622253418,
+      "step": 596
+    },
+    {
+      "epoch": 0.18147275628847176,
+      "grad_norm": 2.334705352783203,
+      "learning_rate": 0.00015,
+      "loss": 4.915990829467773,
+      "step": 597
+    },
+    {
+      "epoch": 0.18177673075461662,
+      "grad_norm": 2.0034000873565674,
+      "learning_rate": 0.00015,
+      "loss": 4.580312728881836,
+      "step": 598
+    },
+    {
+      "epoch": 0.18208070522076145,
+      "grad_norm": 1.9869951009750366,
+      "learning_rate": 0.00015,
+      "loss": 4.7048845291137695,
+      "step": 599
+    },
+    {
+      "epoch": 0.1823846796869063,
+      "grad_norm": 3.051436424255371,
+      "learning_rate": 0.00015,
+      "loss": 4.6412553787231445,
+      "step": 600
+    },
+    {
+      "epoch": 0.18268865415305113,
+      "grad_norm": 2.6076314449310303,
+      "learning_rate": 0.00015,
+      "loss": 4.663086891174316,
+      "step": 601
+    },
+    {
+      "epoch": 0.182992628619196,
+      "grad_norm": 2.3055925369262695,
+      "learning_rate": 0.00015,
+      "loss": 5.0401611328125,
+      "step": 602
+    },
+    {
+      "epoch": 0.18329660308534082,
+      "grad_norm": 2.445986032485962,
+      "learning_rate": 0.00015,
+      "loss": 4.5416436195373535,
+      "step": 603
+    },
+    {
+      "epoch": 0.18360057755148568,
+      "grad_norm": 2.4734063148498535,
+      "learning_rate": 0.00015,
+      "loss": 4.536656379699707,
+      "step": 604
+    },
+    {
+      "epoch": 0.1839045520176305,
+      "grad_norm": 2.341515302658081,
+      "learning_rate": 0.00015,
+      "loss": 4.703493118286133,
+      "step": 605
+    },
+    {
+      "epoch": 0.18420852648377536,
+      "grad_norm": 2.0865321159362793,
+      "learning_rate": 0.00015,
+      "loss": 4.485430717468262,
+      "step": 606
+    },
+    {
+      "epoch": 0.1845125009499202,
+      "grad_norm": 2.0874581336975098,
+      "learning_rate": 0.00015,
+      "loss": 4.5273847579956055,
+      "step": 607
+    },
+    {
+      "epoch": 0.18481647541606505,
+      "grad_norm": 2.6050469875335693,
+      "learning_rate": 0.00015,
+      "loss": 4.6014604568481445,
+      "step": 608
+    },
+    {
+      "epoch": 0.18512044988220988,
+      "grad_norm": 2.7951905727386475,
+      "learning_rate": 0.00015,
+      "loss": 4.670378684997559,
+      "step": 609
+    },
+    {
+      "epoch": 0.18542442434835474,
+      "grad_norm": 2.2661044597625732,
+      "learning_rate": 0.00015,
+      "loss": 4.757679462432861,
+      "step": 610
+    },
+    {
+      "epoch": 0.1857283988144996,
+      "grad_norm": 1.8237788677215576,
+      "learning_rate": 0.00015,
+      "loss": 4.663468360900879,
+      "step": 611
+    },
+    {
+      "epoch": 0.18603237328064443,
+      "grad_norm": 3.7094368934631348,
+      "learning_rate": 0.00015,
+      "loss": 4.616589546203613,
+      "step": 612
+    },
+    {
+      "epoch": 0.18633634774678928,
+      "grad_norm": 3.4019408226013184,
+      "learning_rate": 0.00015,
+      "loss": 4.699448585510254,
+      "step": 613
+    },
+    {
+      "epoch": 0.1866403222129341,
+      "grad_norm": 2.1976773738861084,
+      "learning_rate": 0.00015,
+      "loss": 4.69245719909668,
+      "step": 614
+    },
+    {
+      "epoch": 0.18694429667907897,
+      "grad_norm": 1.9290658235549927,
+      "learning_rate": 0.00015,
+      "loss": 4.545990943908691,
+      "step": 615
+    },
+    {
+      "epoch": 0.1872482711452238,
+      "grad_norm": 2.031838893890381,
+      "learning_rate": 0.00015,
+      "loss": 4.659470558166504,
+      "step": 616
+    },
+    {
+      "epoch": 0.18755224561136866,
+      "grad_norm": 2.029078722000122,
+      "learning_rate": 0.00015,
+      "loss": 4.587153434753418,
+      "step": 617
+    },
+    {
+      "epoch": 0.18785622007751349,
+      "grad_norm": 2.007504940032959,
+      "learning_rate": 0.00015,
+      "loss": 4.500871658325195,
+      "step": 618
+    },
+    {
+      "epoch": 0.18816019454365834,
+      "grad_norm": 2.765791177749634,
+      "learning_rate": 0.00015,
+      "loss": 4.766175270080566,
+      "step": 619
+    },
+    {
+      "epoch": 0.18846416900980317,
+      "grad_norm": 2.2644596099853516,
+      "learning_rate": 0.00015,
+      "loss": 4.543404579162598,
+      "step": 620
+    },
+    {
+      "epoch": 0.18876814347594803,
+      "grad_norm": 2.0469484329223633,
+      "learning_rate": 0.00015,
+      "loss": 4.700861930847168,
+      "step": 621
+    },
+    {
+      "epoch": 0.18907211794209286,
+      "grad_norm": 2.420382022857666,
+      "learning_rate": 0.00015,
+      "loss": 4.71796178817749,
+      "step": 622
+    },
+    {
+      "epoch": 0.18937609240823772,
+      "grad_norm": 2.1186559200286865,
+      "learning_rate": 0.00015,
+      "loss": 4.6682281494140625,
+      "step": 623
+    },
+    {
+      "epoch": 0.18968006687438255,
+      "grad_norm": 2.9782156944274902,
+      "learning_rate": 0.00015,
+      "loss": 4.843927383422852,
+      "step": 624
+    },
+    {
+      "epoch": 0.1899840413405274,
+      "grad_norm": 2.242082357406616,
+      "learning_rate": 0.00015,
+      "loss": 4.791767120361328,
+      "step": 625
+    },
+    {
+      "epoch": 0.19028801580667223,
+      "grad_norm": 2.8726422786712646,
+      "learning_rate": 0.00015,
+      "loss": 4.791931629180908,
+      "step": 626
+    },
+    {
+      "epoch": 0.1905919902728171,
+      "grad_norm": 1.8797760009765625,
+      "learning_rate": 0.00015,
+      "loss": 4.878561019897461,
+      "step": 627
+    },
+    {
+      "epoch": 0.19089596473896192,
+      "grad_norm": 1.8732455968856812,
+      "learning_rate": 0.00015,
+      "loss": 4.830509185791016,
+      "step": 628
+    },
+    {
+      "epoch": 0.19119993920510678,
+      "grad_norm": 1.960268497467041,
+      "learning_rate": 0.00015,
+      "loss": 4.649100303649902,
+      "step": 629
+    },
+    {
+      "epoch": 0.1915039136712516,
+      "grad_norm": 2.010826587677002,
+      "learning_rate": 0.00015,
+      "loss": 4.839844226837158,
+      "step": 630
+    },
+    {
+      "epoch": 0.19180788813739647,
+      "grad_norm": 1.873505711555481,
+      "learning_rate": 0.00015,
+      "loss": 4.720252990722656,
+      "step": 631
+    },
+    {
+      "epoch": 0.1921118626035413,
+      "grad_norm": 2.0552761554718018,
+      "learning_rate": 0.00015,
+      "loss": 4.679355621337891,
+      "step": 632
+    },
+    {
+      "epoch": 0.19241583706968615,
+      "grad_norm": 2.182342290878296,
+      "learning_rate": 0.00015,
+      "loss": 4.929475784301758,
+      "step": 633
+    },
+    {
+      "epoch": 0.19271981153583098,
+      "grad_norm": 2.02223801612854,
+      "learning_rate": 0.00015,
+      "loss": 4.771568298339844,
+      "step": 634
+    },
+    {
+      "epoch": 0.19302378600197584,
+      "grad_norm": 1.8835307359695435,
+      "learning_rate": 0.00015,
+      "loss": 4.635397911071777,
+      "step": 635
+    },
+    {
+      "epoch": 0.19332776046812067,
+      "grad_norm": 1.8479264974594116,
+      "learning_rate": 0.00015,
+      "loss": 4.759753227233887,
+      "step": 636
+    },
+    {
+      "epoch": 0.19363173493426553,
+      "grad_norm": 2.329286813735962,
+      "learning_rate": 0.00015,
+      "loss": 5.154875755310059,
+      "step": 637
+    },
+    {
+      "epoch": 0.19393570940041036,
+      "grad_norm": 2.5284535884857178,
+      "learning_rate": 0.00015,
+      "loss": 4.5214691162109375,
+      "step": 638
+    },
+    {
+      "epoch": 0.1942396838665552,
+      "grad_norm": 3.615879774093628,
+      "learning_rate": 0.00015,
+      "loss": 4.79741096496582,
+      "step": 639
+    },
+    {
+      "epoch": 0.19454365833270004,
+      "grad_norm": 1.9833587408065796,
+      "learning_rate": 0.00015,
+      "loss": 5.069067001342773,
+      "step": 640
+    },
+    {
+      "epoch": 0.1948476327988449,
+      "grad_norm": 2.2228548526763916,
+      "learning_rate": 0.00015,
+      "loss": 4.857247352600098,
+      "step": 641
+    },
+    {
+      "epoch": 0.19515160726498973,
+      "grad_norm": 1.8561897277832031,
+      "learning_rate": 0.00015,
+      "loss": 4.641446113586426,
+      "step": 642
+    },
+    {
+      "epoch": 0.1954555817311346,
+      "grad_norm": 2.030608892440796,
+      "learning_rate": 0.00015,
+      "loss": 4.622519493103027,
+      "step": 643
+    },
+    {
+      "epoch": 0.19575955619727942,
+      "grad_norm": 2.4264798164367676,
+      "learning_rate": 0.00015,
+      "loss": 4.603147029876709,
+      "step": 644
+    },
+    {
+      "epoch": 0.19606353066342427,
+      "grad_norm": 2.12933349609375,
+      "learning_rate": 0.00015,
+      "loss": 4.602532386779785,
+      "step": 645
+    },
+    {
+      "epoch": 0.1963675051295691,
+      "grad_norm": 2.4693524837493896,
+      "learning_rate": 0.00015,
+      "loss": 4.693070411682129,
+      "step": 646
+    },
+    {
+      "epoch": 0.19667147959571396,
+      "grad_norm": 2.2823121547698975,
+      "learning_rate": 0.00015,
+      "loss": 4.587335586547852,
+      "step": 647
+    },
+    {
+      "epoch": 0.1969754540618588,
+      "grad_norm": 1.8455626964569092,
+      "learning_rate": 0.00015,
+      "loss": 4.841438293457031,
+      "step": 648
+    },
+    {
+      "epoch": 0.19727942852800365,
+      "grad_norm": 2.6799581050872803,
+      "learning_rate": 0.00015,
+      "loss": 4.7291579246521,
+      "step": 649
+    },
+    {
+      "epoch": 0.1975834029941485,
+      "grad_norm": 3.15517258644104,
+      "learning_rate": 0.00015,
+      "loss": 4.66948127746582,
+      "step": 650
+    },
+    {
+      "epoch": 0.19788737746029333,
+      "grad_norm": 1.839115023612976,
+      "learning_rate": 0.00015,
+      "loss": 4.564262866973877,
+      "step": 651
+    },
+    {
+      "epoch": 0.1981913519264382,
+      "grad_norm": 2.419416904449463,
+      "learning_rate": 0.00015,
+      "loss": 4.794452667236328,
+      "step": 652
+    },
+    {
+      "epoch": 0.19849532639258302,
+      "grad_norm": 1.9283132553100586,
+      "learning_rate": 0.00015,
+      "loss": 4.6456618309021,
+      "step": 653
+    },
+    {
+      "epoch": 0.19879930085872788,
+      "grad_norm": 1.687606692314148,
+      "learning_rate": 0.00015,
+      "loss": 4.423478603363037,
+      "step": 654
+    },
+    {
+      "epoch": 0.1991032753248727,
+      "grad_norm": 2.002455711364746,
+      "learning_rate": 0.00015,
+      "loss": 4.757028102874756,
+      "step": 655
+    },
+    {
+      "epoch": 0.19940724979101757,
+      "grad_norm": 2.501544952392578,
+      "learning_rate": 0.00015,
+      "loss": 4.741666793823242,
+      "step": 656
+    },
+    {
+      "epoch": 0.1997112242571624,
+      "grad_norm": 1.9457576274871826,
+      "learning_rate": 0.00015,
+      "loss": 4.567790985107422,
+      "step": 657
+    },
+    {
+      "epoch": 0.20001519872330725,
+      "grad_norm": 1.828802466392517,
+      "learning_rate": 0.00015,
+      "loss": 4.823399543762207,
+      "step": 658
+    },
+    {
+      "epoch": 0.20031917318945208,
+      "grad_norm": 2.170285940170288,
+      "learning_rate": 0.00015,
+      "loss": 4.594809532165527,
+      "step": 659
+    },
+    {
+      "epoch": 0.20062314765559694,
+      "grad_norm": 2.4117462635040283,
+      "learning_rate": 0.00015,
+      "loss": 4.487173080444336,
+      "step": 660
+    },
+    {
+      "epoch": 0.20092712212174177,
+      "grad_norm": 2.1024303436279297,
+      "learning_rate": 0.00015,
+      "loss": 4.680421829223633,
+      "step": 661
+    },
+    {
+      "epoch": 0.20123109658788663,
+      "grad_norm": 4.35832405090332,
+      "learning_rate": 0.00015,
+      "loss": 4.335221767425537,
+      "step": 662
+    },
+    {
+      "epoch": 0.20153507105403146,
+      "grad_norm": 2.2309765815734863,
+      "learning_rate": 0.00015,
+      "loss": 4.323690414428711,
+      "step": 663
+    },
+    {
+      "epoch": 0.2018390455201763,
+      "grad_norm": 2.8978116512298584,
+      "learning_rate": 0.00015,
+      "loss": 4.693480491638184,
+      "step": 664
+    },
+    {
+      "epoch": 0.20214301998632114,
+      "grad_norm": 2.0716748237609863,
+      "learning_rate": 0.00015,
+      "loss": 4.545576095581055,
+      "step": 665
+    },
+    {
+      "epoch": 0.202446994452466,
+      "grad_norm": 1.848755955696106,
+      "learning_rate": 0.00015,
+      "loss": 4.524991035461426,
+      "step": 666
+    },
+    {
+      "epoch": 0.20275096891861083,
+      "grad_norm": 1.8698725700378418,
+      "learning_rate": 0.00015,
+      "loss": 4.817290782928467,
+      "step": 667
+    },
+    {
+      "epoch": 0.2030549433847557,
+      "grad_norm": 2.3787505626678467,
+      "learning_rate": 0.00015,
+      "loss": 4.672388553619385,
+      "step": 668
+    },
+    {
+      "epoch": 0.20335891785090052,
+      "grad_norm": 1.937994360923767,
+      "learning_rate": 0.00015,
+      "loss": 4.566434860229492,
+      "step": 669
+    },
+    {
+      "epoch": 0.20366289231704537,
+      "grad_norm": 1.7715823650360107,
+      "learning_rate": 0.00015,
+      "loss": 4.421970367431641,
+      "step": 670
+    },
+    {
+      "epoch": 0.2039668667831902,
+      "grad_norm": 2.468071222305298,
+      "learning_rate": 0.00015,
+      "loss": 4.302404403686523,
+      "step": 671
+    },
+    {
+      "epoch": 0.20427084124933506,
+      "grad_norm": 2.330364942550659,
+      "learning_rate": 0.00015,
+      "loss": 4.659336090087891,
+      "step": 672
+    },
+    {
+      "epoch": 0.2045748157154799,
+      "grad_norm": 2.261566638946533,
+      "learning_rate": 0.00015,
+      "loss": 4.659462928771973,
+      "step": 673
+    },
+    {
+      "epoch": 0.20487879018162475,
+      "grad_norm": 2.0138514041900635,
+      "learning_rate": 0.00015,
+      "loss": 4.539304733276367,
+      "step": 674
+    },
+    {
+      "epoch": 0.20518276464776958,
+      "grad_norm": 2.2087881565093994,
+      "learning_rate": 0.00015,
+      "loss": 4.713778972625732,
+      "step": 675
+    },
+    {
+      "epoch": 0.20548673911391444,
+      "grad_norm": 2.486133337020874,
+      "learning_rate": 0.00015,
+      "loss": 4.888815402984619,
+      "step": 676
+    },
+    {
+      "epoch": 0.20579071358005926,
+      "grad_norm": 1.9386955499649048,
+      "learning_rate": 0.00015,
+      "loss": 4.654658317565918,
+      "step": 677
+    },
+    {
+      "epoch": 0.20609468804620412,
+      "grad_norm": 2.242424726486206,
+      "learning_rate": 0.00015,
+      "loss": 4.878629684448242,
+      "step": 678
+    },
+    {
+      "epoch": 0.20639866251234895,
+      "grad_norm": 2.1355783939361572,
+      "learning_rate": 0.00015,
+      "loss": 4.8423051834106445,
+      "step": 679
+    },
+    {
+      "epoch": 0.2067026369784938,
+      "grad_norm": 2.1109023094177246,
+      "learning_rate": 0.00015,
+      "loss": 4.646300792694092,
+      "step": 680
+    },
+    {
+      "epoch": 0.20700661144463864,
+      "grad_norm": 1.8920774459838867,
+      "learning_rate": 0.00015,
+      "loss": 4.700130462646484,
+      "step": 681
+    },
+    {
+      "epoch": 0.2073105859107835,
+      "grad_norm": 1.764694094657898,
+      "learning_rate": 0.00015,
+      "loss": 4.754611968994141,
+      "step": 682
+    },
+    {
+      "epoch": 0.20761456037692833,
+      "grad_norm": 2.378643035888672,
+      "learning_rate": 0.00015,
+      "loss": 4.656851768493652,
+      "step": 683
+    },
+    {
+      "epoch": 0.20791853484307318,
+      "grad_norm": 2.2966840267181396,
+      "learning_rate": 0.00015,
+      "loss": 4.691801071166992,
+      "step": 684
+    },
+    {
+      "epoch": 0.208222509309218,
+      "grad_norm": 2.7518787384033203,
+      "learning_rate": 0.00015,
+      "loss": 4.549347400665283,
+      "step": 685
+    },
+    {
+      "epoch": 0.20852648377536287,
+      "grad_norm": 2.0341928005218506,
+      "learning_rate": 0.00015,
+      "loss": 4.51059103012085,
+      "step": 686
+    },
+    {
+      "epoch": 0.20883045824150773,
+      "grad_norm": 2.110797643661499,
+      "learning_rate": 0.00015,
+      "loss": 4.519848823547363,
+      "step": 687
+    },
+    {
+      "epoch": 0.20913443270765256,
+      "grad_norm": 2.214777946472168,
+      "learning_rate": 0.00015,
+      "loss": 4.586942195892334,
+      "step": 688
+    },
+    {
+      "epoch": 0.20943840717379741,
+      "grad_norm": 5.06501579284668,
+      "learning_rate": 0.00015,
+      "loss": 5.048894882202148,
+      "step": 689
+    },
+    {
+      "epoch": 0.20974238163994224,
+      "grad_norm": 2.190328598022461,
+      "learning_rate": 0.00015,
+      "loss": 4.999259948730469,
+      "step": 690
+    },
+    {
+      "epoch": 0.2100463561060871,
+      "grad_norm": 1.6447014808654785,
+      "learning_rate": 0.00015,
+      "loss": 4.558644771575928,
+      "step": 691
+    },
+    {
+      "epoch": 0.21035033057223193,
+      "grad_norm": 1.8723785877227783,
+      "learning_rate": 0.00015,
+      "loss": 4.889471530914307,
+      "step": 692
+    },
+    {
+      "epoch": 0.2106543050383768,
+      "grad_norm": 2.342984199523926,
+      "learning_rate": 0.00015,
+      "loss": 4.611535549163818,
+      "step": 693
+    },
+    {
+      "epoch": 0.21095827950452162,
+      "grad_norm": 1.6854853630065918,
+      "learning_rate": 0.00015,
+      "loss": 4.619456768035889,
+      "step": 694
+    },
+    {
+      "epoch": 0.21126225397066647,
+      "grad_norm": 2.175811767578125,
+      "learning_rate": 0.00015,
+      "loss": 4.7791948318481445,
+      "step": 695
+    },
+    {
+      "epoch": 0.2115662284368113,
+      "grad_norm": 2.23468017578125,
+      "learning_rate": 0.00015,
+      "loss": 4.7414231300354,
+      "step": 696
+    },
+    {
+      "epoch": 0.21187020290295616,
+      "grad_norm": 2.133988857269287,
+      "learning_rate": 0.00015,
+      "loss": 4.54448127746582,
+      "step": 697
+    },
+    {
+      "epoch": 0.212174177369101,
+      "grad_norm": 2.5007758140563965,
+      "learning_rate": 0.00015,
+      "loss": 4.7253265380859375,
+      "step": 698
+    },
+    {
+      "epoch": 0.21247815183524585,
+      "grad_norm": 2.6339120864868164,
+      "learning_rate": 0.00015,
+      "loss": 4.656135559082031,
+      "step": 699
+    },
+    {
+      "epoch": 0.21278212630139068,
+      "grad_norm": 1.9814618825912476,
+      "learning_rate": 0.00015,
+      "loss": 4.788064002990723,
+      "step": 700
+    },
+    {
+      "epoch": 0.21308610076753554,
+      "grad_norm": 1.9231650829315186,
+      "learning_rate": 0.00015,
+      "loss": 4.43074893951416,
+      "step": 701
+    },
+    {
+      "epoch": 0.21339007523368037,
+      "grad_norm": 2.3271610736846924,
+      "learning_rate": 0.00015,
+      "loss": 4.431333541870117,
+      "step": 702
+    },
+    {
+      "epoch": 0.21369404969982522,
+      "grad_norm": 1.9826456308364868,
+      "learning_rate": 0.00015,
+      "loss": 4.570574760437012,
+      "step": 703
+    },
+    {
+      "epoch": 0.21399802416597005,
+      "grad_norm": 1.9069780111312866,
+      "learning_rate": 0.00015,
+      "loss": 4.74398136138916,
+      "step": 704
+    },
+    {
+      "epoch": 0.2143019986321149,
+      "grad_norm": 2.70461368560791,
+      "learning_rate": 0.00015,
+      "loss": 4.722800254821777,
+      "step": 705
+    },
+    {
+      "epoch": 0.21460597309825974,
+      "grad_norm": 2.175034523010254,
+      "learning_rate": 0.00015,
+      "loss": 4.627922058105469,
+      "step": 706
+    },
+    {
+      "epoch": 0.2149099475644046,
+      "grad_norm": 1.8454289436340332,
+      "learning_rate": 0.00015,
+      "loss": 4.788971900939941,
+      "step": 707
+    },
+    {
+      "epoch": 0.21521392203054943,
+      "grad_norm": 2.56282377243042,
+      "learning_rate": 0.00015,
+      "loss": 4.8575544357299805,
+      "step": 708
+    },
+    {
+      "epoch": 0.21551789649669428,
+      "grad_norm": 1.8437303304672241,
+      "learning_rate": 0.00015,
+      "loss": 4.325046539306641,
+      "step": 709
+    },
+    {
+      "epoch": 0.2158218709628391,
+      "grad_norm": 1.9878103733062744,
+      "learning_rate": 0.00015,
+      "loss": 4.656960487365723,
+      "step": 710
+    },
+    {
+      "epoch": 0.21612584542898397,
+      "grad_norm": 1.6025094985961914,
+      "learning_rate": 0.00015,
+      "loss": 4.499412536621094,
+      "step": 711
+    },
+    {
+      "epoch": 0.2164298198951288,
+      "grad_norm": 1.8254647254943848,
+      "learning_rate": 0.00015,
+      "loss": 4.562152862548828,
+      "step": 712
+    },
+    {
+      "epoch": 0.21673379436127366,
+      "grad_norm": 2.0381624698638916,
+      "learning_rate": 0.00015,
+      "loss": 4.707602024078369,
+      "step": 713
+    },
+    {
+      "epoch": 0.2170377688274185,
+      "grad_norm": 2.4426968097686768,
+      "learning_rate": 0.00015,
+      "loss": 4.450949192047119,
+      "step": 714
+    },
+    {
+      "epoch": 0.21734174329356334,
+      "grad_norm": 2.4077508449554443,
+      "learning_rate": 0.00015,
+      "loss": 4.893795013427734,
+      "step": 715
+    },
+    {
+      "epoch": 0.21764571775970817,
+      "grad_norm": 2.425611734390259,
+      "learning_rate": 0.00015,
+      "loss": 4.89498233795166,
+      "step": 716
+    },
+    {
+      "epoch": 0.21794969222585303,
+      "grad_norm": 2.4420552253723145,
+      "learning_rate": 0.00015,
+      "loss": 4.518195152282715,
+      "step": 717
+    },
+    {
+      "epoch": 0.21825366669199786,
+      "grad_norm": 2.568052291870117,
+      "learning_rate": 0.00015,
+      "loss": 4.682669162750244,
+      "step": 718
+    },
+    {
+      "epoch": 0.21855764115814272,
+      "grad_norm": 2.586857318878174,
+      "learning_rate": 0.00015,
+      "loss": 4.624917507171631,
+      "step": 719
+    },
+    {
+      "epoch": 0.21886161562428755,
+      "grad_norm": 1.996894359588623,
+      "learning_rate": 0.00015,
+      "loss": 4.492257118225098,
+      "step": 720
+    },
+    {
+      "epoch": 0.2191655900904324,
+      "grad_norm": 1.8800965547561646,
+      "learning_rate": 0.00015,
+      "loss": 4.423683166503906,
+      "step": 721
+    },
+    {
+      "epoch": 0.21946956455657723,
+      "grad_norm": 2.1609301567077637,
+      "learning_rate": 0.00015,
+      "loss": 4.780789375305176,
+      "step": 722
+    },
+    {
+      "epoch": 0.2197735390227221,
+      "grad_norm": 3.237112283706665,
+      "learning_rate": 0.00015,
+      "loss": 4.381412506103516,
+      "step": 723
+    },
+    {
+      "epoch": 0.22007751348886692,
+      "grad_norm": 2.1442668437957764,
+      "learning_rate": 0.00015,
+      "loss": 4.7866950035095215,
+      "step": 724
+    },
+    {
+      "epoch": 0.22038148795501178,
+      "grad_norm": 2.355821371078491,
+      "learning_rate": 0.00015,
+      "loss": 4.416647434234619,
+      "step": 725
+    },
+    {
+      "epoch": 0.22068546242115664,
+      "grad_norm": 2.134929895401001,
+      "learning_rate": 0.00015,
+      "loss": 4.61026668548584,
+      "step": 726
+    },
+    {
+      "epoch": 0.22098943688730147,
+      "grad_norm": 2.431623697280884,
+      "learning_rate": 0.00015,
+      "loss": 4.348535537719727,
+      "step": 727
+    },
+    {
+      "epoch": 0.22129341135344632,
+      "grad_norm": 2.2406973838806152,
+      "learning_rate": 0.00015,
+      "loss": 4.695709705352783,
+      "step": 728
+    },
+    {
+      "epoch": 0.22159738581959115,
+      "grad_norm": 1.8269048929214478,
+      "learning_rate": 0.00015,
+      "loss": 4.750170707702637,
+      "step": 729
+    },
+    {
+      "epoch": 0.221901360285736,
+      "grad_norm": 1.5366374254226685,
+      "learning_rate": 0.00015,
+      "loss": 4.794458389282227,
+      "step": 730
+    },
+    {
+      "epoch": 0.22220533475188084,
+      "grad_norm": 2.476262092590332,
+      "learning_rate": 0.00015,
+      "loss": 4.608065605163574,
+      "step": 731
+    },
+    {
+      "epoch": 0.2225093092180257,
+      "grad_norm": 2.2089955806732178,
+      "learning_rate": 0.00015,
+      "loss": 4.954866409301758,
+      "step": 732
+    },
+    {
+      "epoch": 0.22281328368417053,
+      "grad_norm": 1.9864741563796997,
+      "learning_rate": 0.00015,
+      "loss": 4.482414722442627,
+      "step": 733
+    },
+    {
+      "epoch": 0.22311725815031538,
+      "grad_norm": 1.6228045225143433,
+      "learning_rate": 0.00015,
+      "loss": 4.671963691711426,
+      "step": 734
+    },
+    {
+      "epoch": 0.2234212326164602,
+      "grad_norm": 1.6276811361312866,
+      "learning_rate": 0.00015,
+      "loss": 4.604326248168945,
+      "step": 735
+    },
+    {
+      "epoch": 0.22372520708260507,
+      "grad_norm": 1.8677018880844116,
+      "learning_rate": 0.00015,
+      "loss": 4.223974227905273,
+      "step": 736
+    },
+    {
+      "epoch": 0.2240291815487499,
+      "grad_norm": 1.9835920333862305,
+      "learning_rate": 0.00015,
+      "loss": 4.56862735748291,
+      "step": 737
+    },
+    {
+      "epoch": 0.22433315601489476,
+      "grad_norm": 1.8421988487243652,
+      "learning_rate": 0.00015,
+      "loss": 4.78159236907959,
+      "step": 738
+    },
+    {
+      "epoch": 0.2246371304810396,
+      "grad_norm": 2.1905436515808105,
+      "learning_rate": 0.00015,
+      "loss": 4.536891937255859,
+      "step": 739
+    },
+    {
+      "epoch": 0.22494110494718444,
+      "grad_norm": 2.221676826477051,
+      "learning_rate": 0.00015,
+      "loss": 4.850802421569824,
+      "step": 740
+    },
+    {
+      "epoch": 0.22524507941332927,
+      "grad_norm": 1.8638346195220947,
+      "learning_rate": 0.00015,
+      "loss": 4.691383361816406,
+      "step": 741
+    },
+    {
+      "epoch": 0.22554905387947413,
+      "grad_norm": 2.134999990463257,
+      "learning_rate": 0.00015,
+      "loss": 4.783224582672119,
+      "step": 742
+    },
+    {
+      "epoch": 0.22585302834561896,
+      "grad_norm": 2.2665534019470215,
+      "learning_rate": 0.00015,
+      "loss": 4.699117660522461,
+      "step": 743
+    },
+    {
+      "epoch": 0.22615700281176382,
+      "grad_norm": 2.204928398132324,
+      "learning_rate": 0.00015,
+      "loss": 4.539507865905762,
+      "step": 744
+    },
+    {
+      "epoch": 0.22646097727790865,
+      "grad_norm": 1.822850227355957,
+      "learning_rate": 0.00015,
+      "loss": 4.58991813659668,
+      "step": 745
+    },
+    {
+      "epoch": 0.2267649517440535,
+      "grad_norm": 2.519963026046753,
+      "learning_rate": 0.00015,
+      "loss": 4.205777168273926,
+      "step": 746
+    },
+    {
+      "epoch": 0.22706892621019834,
+      "grad_norm": 2.604090929031372,
+      "learning_rate": 0.00015,
+      "loss": 4.801619529724121,
+      "step": 747
+    },
+    {
+      "epoch": 0.2273729006763432,
+      "grad_norm": 2.07092547416687,
+      "learning_rate": 0.00015,
+      "loss": 4.666153907775879,
+      "step": 748
+    },
+    {
+      "epoch": 0.22767687514248802,
+      "grad_norm": 1.877859115600586,
+      "learning_rate": 0.00015,
+      "loss": 4.494083404541016,
+      "step": 749
+    },
+    {
+      "epoch": 0.22798084960863288,
+      "grad_norm": 2.2293810844421387,
+      "learning_rate": 0.00015,
+      "loss": 4.5393452644348145,
+      "step": 750
+    },
+    {
+      "epoch": 0.2282848240747777,
+      "grad_norm": 2.426513195037842,
+      "learning_rate": 0.00015,
+      "loss": 4.496891498565674,
+      "step": 751
+    },
+    {
+      "epoch": 0.22858879854092257,
+      "grad_norm": 2.1946427822113037,
+      "learning_rate": 0.00015,
+      "loss": 4.455080509185791,
+      "step": 752
+    },
+    {
+      "epoch": 0.2288927730070674,
+      "grad_norm": 1.9524500370025635,
+      "learning_rate": 0.00015,
+      "loss": 4.627020835876465,
+      "step": 753
+    },
+    {
+      "epoch": 0.22919674747321225,
+      "grad_norm": 1.9506654739379883,
+      "learning_rate": 0.00015,
+      "loss": 4.83452033996582,
+      "step": 754
+    },
+    {
+      "epoch": 0.22950072193935708,
+      "grad_norm": 2.0084433555603027,
+      "learning_rate": 0.00015,
+      "loss": 4.850553512573242,
+      "step": 755
+    },
+    {
+      "epoch": 0.22980469640550194,
+      "grad_norm": 1.8148285150527954,
+      "learning_rate": 0.00015,
+      "loss": 5.001620769500732,
+      "step": 756
+    },
+    {
+      "epoch": 0.23010867087164677,
+      "grad_norm": 2.0593225955963135,
+      "learning_rate": 0.00015,
+      "loss": 4.584716796875,
+      "step": 757
+    },
+    {
+      "epoch": 0.23041264533779163,
+      "grad_norm": 2.20324969291687,
+      "learning_rate": 0.00015,
+      "loss": 4.362469673156738,
+      "step": 758
+    },
+    {
+      "epoch": 0.23071661980393646,
+      "grad_norm": 2.4480361938476562,
+      "learning_rate": 0.00015,
+      "loss": 4.557003021240234,
+      "step": 759
+    },
+    {
+      "epoch": 0.23102059427008131,
+      "grad_norm": 1.8594508171081543,
+      "learning_rate": 0.00015,
+      "loss": 4.4106597900390625,
+      "step": 760
+    },
+    {
+      "epoch": 0.23132456873622614,
+      "grad_norm": 1.9597049951553345,
+      "learning_rate": 0.00015,
+      "loss": 4.504006385803223,
+      "step": 761
+    },
+    {
+      "epoch": 0.231628543202371,
+      "grad_norm": 2.1989340782165527,
+      "learning_rate": 0.00015,
+      "loss": 4.612226486206055,
+      "step": 762
+    },
+    {
+      "epoch": 0.23193251766851586,
+      "grad_norm": 2.219116449356079,
+      "learning_rate": 0.00015,
+      "loss": 4.418520927429199,
+      "step": 763
+    },
+    {
+      "epoch": 0.2322364921346607,
+      "grad_norm": 3.2968645095825195,
+      "learning_rate": 0.00015,
+      "loss": 4.364174842834473,
+      "step": 764
+    },
+    {
+      "epoch": 0.23254046660080555,
+      "grad_norm": 2.012354850769043,
+      "learning_rate": 0.00015,
+      "loss": 4.333540439605713,
+      "step": 765
+    },
+    {
+      "epoch": 0.23284444106695038,
+      "grad_norm": 2.231611967086792,
+      "learning_rate": 0.00015,
+      "loss": 4.65167236328125,
+      "step": 766
+    },
+    {
+      "epoch": 0.23314841553309523,
+      "grad_norm": 1.8087263107299805,
+      "learning_rate": 0.00015,
+      "loss": 4.595836162567139,
+      "step": 767
+    },
+    {
+      "epoch": 0.23345238999924006,
+      "grad_norm": 1.675134301185608,
+      "learning_rate": 0.00015,
+      "loss": 4.539973258972168,
+      "step": 768
+    },
+    {
+      "epoch": 0.23375636446538492,
+      "grad_norm": 2.2017133235931396,
+      "learning_rate": 0.00015,
+      "loss": 4.697962760925293,
+      "step": 769
+    },
+    {
+      "epoch": 0.23406033893152975,
+      "grad_norm": 2.8612241744995117,
+      "learning_rate": 0.00015,
+      "loss": 4.472855567932129,
+      "step": 770
+    },
+    {
+      "epoch": 0.2343643133976746,
+      "grad_norm": 1.8407312631607056,
+      "learning_rate": 0.00015,
+      "loss": 4.421652793884277,
+      "step": 771
+    },
+    {
+      "epoch": 0.23466828786381944,
+      "grad_norm": 2.081300973892212,
+      "learning_rate": 0.00015,
+      "loss": 4.454488754272461,
+      "step": 772
+    },
+    {
+      "epoch": 0.2349722623299643,
+      "grad_norm": 2.280085563659668,
+      "learning_rate": 0.00015,
+      "loss": 4.646749496459961,
+      "step": 773
+    },
+    {
+      "epoch": 0.23527623679610912,
+      "grad_norm": 1.7693480253219604,
+      "learning_rate": 0.00015,
+      "loss": 4.51268196105957,
+      "step": 774
+    },
+    {
+      "epoch": 0.23558021126225398,
+      "grad_norm": 2.299192428588867,
+      "learning_rate": 0.00015,
+      "loss": 4.732900619506836,
+      "step": 775
+    },
+    {
+      "epoch": 0.2358841857283988,
+      "grad_norm": 2.0201873779296875,
+      "learning_rate": 0.00015,
+      "loss": 4.193674087524414,
+      "step": 776
+    },
+    {
+      "epoch": 0.23618816019454367,
+      "grad_norm": 1.995118260383606,
+      "learning_rate": 0.00015,
+      "loss": 4.59087610244751,
+      "step": 777
+    },
+    {
+      "epoch": 0.2364921346606885,
+      "grad_norm": 1.807499885559082,
+      "learning_rate": 0.00015,
+      "loss": 4.579825401306152,
+      "step": 778
+    },
+    {
+      "epoch": 0.23679610912683335,
+      "grad_norm": 1.647511601448059,
+      "learning_rate": 0.00015,
+      "loss": 4.270777702331543,
+      "step": 779
+    },
+    {
+      "epoch": 0.23710008359297818,
+      "grad_norm": 1.8517082929611206,
+      "learning_rate": 0.00015,
+      "loss": 4.662494659423828,
+      "step": 780
+    },
+    {
+      "epoch": 0.23740405805912304,
+      "grad_norm": 2.0175108909606934,
+      "learning_rate": 0.00015,
+      "loss": 4.42941951751709,
+      "step": 781
+    },
+    {
+      "epoch": 0.23770803252526787,
+      "grad_norm": 2.321638822555542,
+      "learning_rate": 0.00015,
+      "loss": 4.445045471191406,
+      "step": 782
+    },
+    {
+      "epoch": 0.23801200699141273,
+      "grad_norm": 2.969998598098755,
+      "learning_rate": 0.00015,
+      "loss": 4.479081630706787,
+      "step": 783
+    },
+    {
+      "epoch": 0.23831598145755756,
+      "grad_norm": 2.2516160011291504,
+      "learning_rate": 0.00015,
+      "loss": 4.505674362182617,
+      "step": 784
+    },
+    {
+      "epoch": 0.23861995592370241,
+      "grad_norm": 2.144613742828369,
+      "learning_rate": 0.00015,
+      "loss": 4.600666046142578,
+      "step": 785
+    },
+    {
+      "epoch": 0.23892393038984724,
+      "grad_norm": 2.7086849212646484,
+      "learning_rate": 0.00015,
+      "loss": 4.816709518432617,
+      "step": 786
+    },
+    {
+      "epoch": 0.2392279048559921,
+      "grad_norm": 1.910455346107483,
+      "learning_rate": 0.00015,
+      "loss": 4.776810646057129,
+      "step": 787
+    },
+    {
+      "epoch": 0.23953187932213693,
+      "grad_norm": 2.18768048286438,
+      "learning_rate": 0.00015,
+      "loss": 4.454985618591309,
+      "step": 788
+    },
+    {
+      "epoch": 0.2398358537882818,
+      "grad_norm": 2.0085535049438477,
+      "learning_rate": 0.00015,
+      "loss": 4.738059043884277,
+      "step": 789
+    },
+    {
+      "epoch": 0.24013982825442662,
+      "grad_norm": 2.0622527599334717,
+      "learning_rate": 0.00015,
+      "loss": 4.484880447387695,
+      "step": 790
+    },
+    {
+      "epoch": 0.24044380272057148,
+      "grad_norm": 2.048443078994751,
+      "learning_rate": 0.00015,
+      "loss": 4.668357849121094,
+      "step": 791
+    },
+    {
+      "epoch": 0.2407477771867163,
+      "grad_norm": 1.8799495697021484,
+      "learning_rate": 0.00015,
+      "loss": 4.351439952850342,
+      "step": 792
+    },
+    {
+      "epoch": 0.24105175165286116,
+      "grad_norm": 2.185328960418701,
+      "learning_rate": 0.00015,
+      "loss": 4.785930633544922,
+      "step": 793
+    },
+    {
+      "epoch": 0.241355726119006,
+      "grad_norm": 2.967383861541748,
+      "learning_rate": 0.00015,
+      "loss": 4.653923034667969,
+      "step": 794
+    },
+    {
+      "epoch": 0.24165970058515085,
+      "grad_norm": 1.8684638738632202,
+      "learning_rate": 0.00015,
+      "loss": 4.640868186950684,
+      "step": 795
+    },
+    {
+      "epoch": 0.24196367505129568,
+      "grad_norm": 2.07366943359375,
+      "learning_rate": 0.00015,
+      "loss": 4.50937557220459,
+      "step": 796
+    },
+    {
+      "epoch": 0.24226764951744054,
+      "grad_norm": 2.0127851963043213,
+      "learning_rate": 0.00015,
+      "loss": 4.627388000488281,
+      "step": 797
+    },
+    {
+      "epoch": 0.24257162398358537,
+      "grad_norm": 2.086991786956787,
+      "learning_rate": 0.00015,
+      "loss": 4.555594444274902,
+      "step": 798
+    },
+    {
+      "epoch": 0.24287559844973022,
+      "grad_norm": 2.3385019302368164,
+      "learning_rate": 0.00015,
+      "loss": 4.706195831298828,
+      "step": 799
+    },
+    {
+      "epoch": 0.24317957291587505,
+      "grad_norm": 2.315903902053833,
+      "learning_rate": 0.00015,
+      "loss": 4.841514587402344,
+      "step": 800
+    },
+    {
+      "epoch": 0.2434835473820199,
+      "grad_norm": 2.2510430812835693,
+      "learning_rate": 0.00015,
+      "loss": 4.410776138305664,
+      "step": 801
+    },
+    {
+      "epoch": 0.24378752184816477,
+      "grad_norm": 2.4691200256347656,
+      "learning_rate": 0.00015,
+      "loss": 4.663663864135742,
+      "step": 802
+    },
+    {
+      "epoch": 0.2440914963143096,
+      "grad_norm": 1.9908934831619263,
+      "learning_rate": 0.00015,
+      "loss": 4.563068389892578,
+      "step": 803
+    },
+    {
+      "epoch": 0.24439547078045445,
+      "grad_norm": 4.846072673797607,
+      "learning_rate": 0.00015,
+      "loss": 4.771219253540039,
+      "step": 804
+    },
+    {
+      "epoch": 0.24469944524659928,
+      "grad_norm": 2.0190696716308594,
+      "learning_rate": 0.00015,
+      "loss": 4.472007751464844,
+      "step": 805
+    },
+    {
+      "epoch": 0.24500341971274414,
+      "grad_norm": 2.0539357662200928,
+      "learning_rate": 0.00015,
+      "loss": 4.275585174560547,
+      "step": 806
+    },
+    {
+      "epoch": 0.24530739417888897,
+      "grad_norm": 1.9895809888839722,
+      "learning_rate": 0.00015,
+      "loss": 4.184213638305664,
+      "step": 807
+    },
+    {
+      "epoch": 0.24561136864503383,
+      "grad_norm": 1.9838426113128662,
+      "learning_rate": 0.00015,
+      "loss": 4.286787986755371,
+      "step": 808
+    },
+    {
+      "epoch": 0.24591534311117866,
+      "grad_norm": 2.125953197479248,
+      "learning_rate": 0.00015,
+      "loss": 4.642009735107422,
+      "step": 809
+    },
+    {
+      "epoch": 0.24621931757732352,
+      "grad_norm": 1.680031657218933,
+      "learning_rate": 0.00015,
+      "loss": 4.608014106750488,
+      "step": 810
+    },
+    {
+      "epoch": 0.24652329204346835,
+      "grad_norm": 2.430569887161255,
+      "learning_rate": 0.00015,
+      "loss": 4.580404758453369,
+      "step": 811
+    },
+    {
+      "epoch": 0.2468272665096132,
+      "grad_norm": 2.2124640941619873,
+      "learning_rate": 0.00015,
+      "loss": 4.710816383361816,
+      "step": 812
+    },
+    {
+      "epoch": 0.24713124097575803,
+      "grad_norm": 2.66298770904541,
+      "learning_rate": 0.00015,
+      "loss": 4.631195068359375,
+      "step": 813
+    },
+    {
+      "epoch": 0.2474352154419029,
+      "grad_norm": 1.9879980087280273,
+      "learning_rate": 0.00015,
+      "loss": 4.49324893951416,
+      "step": 814
+    },
+    {
+      "epoch": 0.24773918990804772,
+      "grad_norm": 1.845962643623352,
+      "learning_rate": 0.00015,
+      "loss": 4.3389434814453125,
+      "step": 815
+    },
+    {
+      "epoch": 0.24804316437419258,
+      "grad_norm": 1.9998226165771484,
+      "learning_rate": 0.00015,
+      "loss": 4.476945877075195,
+      "step": 816
+    },
+    {
+      "epoch": 0.2483471388403374,
+      "grad_norm": 1.9951765537261963,
+      "learning_rate": 0.00015,
+      "loss": 4.289376258850098,
+      "step": 817
+    },
+    {
+      "epoch": 0.24865111330648226,
+      "grad_norm": 2.10616397857666,
+      "learning_rate": 0.00015,
+      "loss": 4.4439616203308105,
+      "step": 818
+    },
+    {
+      "epoch": 0.2489550877726271,
+      "grad_norm": 1.8434419631958008,
+      "learning_rate": 0.00015,
+      "loss": 4.695892333984375,
+      "step": 819
+    },
+    {
+      "epoch": 0.24925906223877195,
+      "grad_norm": 2.3386635780334473,
+      "learning_rate": 0.00015,
+      "loss": 4.792157173156738,
+      "step": 820
+    },
+    {
+      "epoch": 0.24956303670491678,
+      "grad_norm": 2.262343168258667,
+      "learning_rate": 0.00015,
+      "loss": 4.801372051239014,
+      "step": 821
+    },
+    {
+      "epoch": 0.24986701117106164,
+      "grad_norm": 2.120734930038452,
+      "learning_rate": 0.00015,
+      "loss": 4.629687309265137,
+      "step": 822
+    },
+    {
+      "epoch": 0.25017098563720647,
+      "grad_norm": 1.8164788484573364,
+      "learning_rate": 0.00015,
+      "loss": 4.419816970825195,
+      "step": 823
+    },
+    {
+      "epoch": 0.2504749601033513,
+      "grad_norm": 2.6520559787750244,
+      "learning_rate": 0.00015,
+      "loss": 4.42047119140625,
+      "step": 824
+    },
+    {
+      "epoch": 0.2507789345694962,
+      "grad_norm": 2.4225847721099854,
+      "learning_rate": 0.00015,
+      "loss": 4.523772716522217,
+      "step": 825
+    },
+    {
+      "epoch": 0.251082909035641,
+      "grad_norm": 2.0425264835357666,
+      "learning_rate": 0.00015,
+      "loss": 4.5266499519348145,
+      "step": 826
+    },
+    {
+      "epoch": 0.25138688350178584,
+      "grad_norm": 2.0179030895233154,
+      "learning_rate": 0.00015,
+      "loss": 4.52508544921875,
+      "step": 827
+    },
+    {
+      "epoch": 0.25169085796793067,
+      "grad_norm": 2.0066308975219727,
+      "learning_rate": 0.00015,
+      "loss": 4.336664199829102,
+      "step": 828
+    },
+    {
+      "epoch": 0.25199483243407556,
+      "grad_norm": 2.422898292541504,
+      "learning_rate": 0.00015,
+      "loss": 4.2602972984313965,
+      "step": 829
+    },
+    {
+      "epoch": 0.2522988069002204,
+      "grad_norm": 1.803850769996643,
+      "learning_rate": 0.00015,
+      "loss": 4.267040729522705,
+      "step": 830
+    },
+    {
+      "epoch": 0.2526027813663652,
+      "grad_norm": 1.8568986654281616,
+      "learning_rate": 0.00015,
+      "loss": 4.399720191955566,
+      "step": 831
+    },
+    {
+      "epoch": 0.25290675583251004,
+      "grad_norm": 1.9580368995666504,
+      "learning_rate": 0.00015,
+      "loss": 4.297675609588623,
+      "step": 832
+    },
+    {
+      "epoch": 0.25321073029865493,
+      "grad_norm": 2.1572461128234863,
+      "learning_rate": 0.00015,
+      "loss": 4.405695915222168,
+      "step": 833
+    },
+    {
+      "epoch": 0.25351470476479976,
+      "grad_norm": 1.6348316669464111,
+      "learning_rate": 0.00015,
+      "loss": 4.534473896026611,
+      "step": 834
+    },
+    {
+      "epoch": 0.2538186792309446,
+      "grad_norm": 2.7165610790252686,
+      "learning_rate": 0.00015,
+      "loss": 4.367175102233887,
+      "step": 835
+    },
+    {
+      "epoch": 0.2541226536970894,
+      "grad_norm": 1.6767299175262451,
+      "learning_rate": 0.00015,
+      "loss": 4.516571521759033,
+      "step": 836
+    },
+    {
+      "epoch": 0.2544266281632343,
+      "grad_norm": 2.2659075260162354,
+      "learning_rate": 0.00015,
+      "loss": 4.286858558654785,
+      "step": 837
+    },
+    {
+      "epoch": 0.25473060262937913,
+      "grad_norm": 1.8916198015213013,
+      "learning_rate": 0.00015,
+      "loss": 4.520571708679199,
+      "step": 838
+    },
+    {
+      "epoch": 0.25503457709552396,
+      "grad_norm": 2.013199806213379,
+      "learning_rate": 0.00015,
+      "loss": 4.64973258972168,
+      "step": 839
+    },
+    {
+      "epoch": 0.25533855156166885,
+      "grad_norm": 1.9138273000717163,
+      "learning_rate": 0.00015,
+      "loss": 4.8362836837768555,
+      "step": 840
+    },
+    {
+      "epoch": 0.2556425260278137,
+      "grad_norm": 2.064126491546631,
+      "learning_rate": 0.00015,
+      "loss": 4.735012054443359,
+      "step": 841
+    },
+    {
+      "epoch": 0.2559465004939585,
+      "grad_norm": 2.1910362243652344,
+      "learning_rate": 0.00015,
+      "loss": 4.524472236633301,
+      "step": 842
+    },
+    {
+      "epoch": 0.25625047496010334,
+      "grad_norm": 2.3203794956207275,
+      "learning_rate": 0.00015,
+      "loss": 4.502038955688477,
+      "step": 843
+    },
+    {
+      "epoch": 0.2565544494262482,
+      "grad_norm": 2.5984344482421875,
+      "learning_rate": 0.00015,
+      "loss": 4.40181827545166,
+      "step": 844
+    },
+    {
+      "epoch": 0.25685842389239305,
+      "grad_norm": 1.8128288984298706,
+      "learning_rate": 0.00015,
+      "loss": 4.661513328552246,
+      "step": 845
+    },
+    {
+      "epoch": 0.2571623983585379,
+      "grad_norm": 1.8038901090621948,
+      "learning_rate": 0.00015,
+      "loss": 4.289132118225098,
+      "step": 846
+    },
+    {
+      "epoch": 0.2574663728246827,
+      "grad_norm": 1.9731134176254272,
+      "learning_rate": 0.00015,
+      "loss": 4.598901748657227,
+      "step": 847
+    },
+    {
+      "epoch": 0.2577703472908276,
+      "grad_norm": 2.0627293586730957,
+      "learning_rate": 0.00015,
+      "loss": 4.4145708084106445,
+      "step": 848
+    },
+    {
+      "epoch": 0.2580743217569724,
+      "grad_norm": 1.7435401678085327,
+      "learning_rate": 0.00015,
+      "loss": 4.396785736083984,
+      "step": 849
+    },
+    {
+      "epoch": 0.25837829622311725,
+      "grad_norm": 1.9624364376068115,
+      "learning_rate": 0.00015,
+      "loss": 4.230037212371826,
+      "step": 850
+    },
+    {
+      "epoch": 0.2586822706892621,
+      "grad_norm": 4.1804938316345215,
+      "learning_rate": 0.00015,
+      "loss": 4.528890609741211,
+      "step": 851
+    },
+    {
+      "epoch": 0.25898624515540697,
+      "grad_norm": 2.1737167835235596,
+      "learning_rate": 0.00015,
+      "loss": 4.343877792358398,
+      "step": 852
+    },
+    {
+      "epoch": 0.2592902196215518,
+      "grad_norm": 3.161729097366333,
+      "learning_rate": 0.00015,
+      "loss": 4.617306709289551,
+      "step": 853
+    },
+    {
+      "epoch": 0.25959419408769663,
+      "grad_norm": 1.8518584966659546,
+      "learning_rate": 0.00015,
+      "loss": 4.3760881423950195,
+      "step": 854
+    },
+    {
+      "epoch": 0.25989816855384146,
+      "grad_norm": 1.8959300518035889,
+      "learning_rate": 0.00015,
+      "loss": 4.863279819488525,
+      "step": 855
+    },
+    {
+      "epoch": 0.26020214301998634,
+      "grad_norm": 2.2699992656707764,
+      "learning_rate": 0.00015,
+      "loss": 4.236268043518066,
+      "step": 856
+    },
+    {
+      "epoch": 0.2605061174861312,
+      "grad_norm": 1.9495069980621338,
+      "learning_rate": 0.00015,
+      "loss": 4.660202503204346,
+      "step": 857
+    },
+    {
+      "epoch": 0.260810091952276,
+      "grad_norm": 3.1187398433685303,
+      "learning_rate": 0.00015,
+      "loss": 4.455247402191162,
+      "step": 858
+    },
+    {
+      "epoch": 0.26111406641842083,
+      "grad_norm": 1.9775768518447876,
+      "learning_rate": 0.00015,
+      "loss": 4.348986625671387,
+      "step": 859
+    },
+    {
+      "epoch": 0.2614180408845657,
+      "grad_norm": 8.115832328796387,
+      "learning_rate": 0.00015,
+      "loss": 5.061985015869141,
+      "step": 860
+    },
+    {
+      "epoch": 0.26172201535071055,
+      "grad_norm": 2.765378475189209,
+      "learning_rate": 0.00015,
+      "loss": 4.374677658081055,
+      "step": 861
+    },
+    {
+      "epoch": 0.2620259898168554,
+      "grad_norm": 2.0354931354522705,
+      "learning_rate": 0.00015,
+      "loss": 4.587867736816406,
+      "step": 862
+    },
+    {
+      "epoch": 0.2623299642830002,
+      "grad_norm": 2.122919797897339,
+      "learning_rate": 0.00015,
+      "loss": 4.362419128417969,
+      "step": 863
+    },
+    {
+      "epoch": 0.2626339387491451,
+      "grad_norm": 2.4674181938171387,
+      "learning_rate": 0.00015,
+      "loss": 4.0872111320495605,
+      "step": 864
+    },
+    {
+      "epoch": 0.2629379132152899,
+      "grad_norm": 2.413919687271118,
+      "learning_rate": 0.00015,
+      "loss": 4.683576583862305,
+      "step": 865
+    },
+    {
+      "epoch": 0.26324188768143475,
+      "grad_norm": 2.811927318572998,
+      "learning_rate": 0.00015,
+      "loss": 4.536355018615723,
+      "step": 866
+    },
+    {
+      "epoch": 0.2635458621475796,
+      "grad_norm": 2.0893516540527344,
+      "learning_rate": 0.00015,
+      "loss": 4.738877773284912,
+      "step": 867
+    },
+    {
+      "epoch": 0.26384983661372446,
+      "grad_norm": 2.0967483520507812,
+      "learning_rate": 0.00015,
+      "loss": 4.520099639892578,
+      "step": 868
+    },
+    {
+      "epoch": 0.2641538110798693,
+      "grad_norm": 2.6684787273406982,
+      "learning_rate": 0.00015,
+      "loss": 4.398592948913574,
+      "step": 869
+    },
+    {
+      "epoch": 0.2644577855460141,
+      "grad_norm": 1.8888270854949951,
+      "learning_rate": 0.00015,
+      "loss": 5.060788631439209,
+      "step": 870
+    },
+    {
+      "epoch": 0.26476176001215895,
+      "grad_norm": 1.9709162712097168,
+      "learning_rate": 0.00015,
+      "loss": 4.547727108001709,
+      "step": 871
+    },
+    {
+      "epoch": 0.26506573447830384,
+      "grad_norm": 2.6234631538391113,
+      "learning_rate": 0.00015,
+      "loss": 4.301576614379883,
+      "step": 872
+    },
+    {
+      "epoch": 0.26536970894444867,
+      "grad_norm": 2.723207712173462,
+      "learning_rate": 0.00015,
+      "loss": 4.397118091583252,
+      "step": 873
+    },
+    {
+      "epoch": 0.2656736834105935,
+      "grad_norm": 1.8279407024383545,
+      "learning_rate": 0.00015,
+      "loss": 4.582146644592285,
+      "step": 874
+    },
+    {
+      "epoch": 0.2659776578767383,
+      "grad_norm": 2.235320806503296,
+      "learning_rate": 0.00015,
+      "loss": 4.565396308898926,
+      "step": 875
+    },
+    {
+      "epoch": 0.2662816323428832,
+      "grad_norm": 1.8793033361434937,
+      "learning_rate": 0.00015,
+      "loss": 4.531579971313477,
+      "step": 876
+    },
+    {
+      "epoch": 0.26658560680902804,
+      "grad_norm": 2.5706987380981445,
+      "learning_rate": 0.00015,
+      "loss": 4.648132801055908,
+      "step": 877
+    },
+    {
+      "epoch": 0.26688958127517287,
+      "grad_norm": 2.3678460121154785,
+      "learning_rate": 0.00015,
+      "loss": 4.120333671569824,
+      "step": 878
+    },
+    {
+      "epoch": 0.26719355574131776,
+      "grad_norm": 1.7416727542877197,
+      "learning_rate": 0.00015,
+      "loss": 4.3199028968811035,
+      "step": 879
+    },
+    {
+      "epoch": 0.2674975302074626,
+      "grad_norm": 2.417978048324585,
+      "learning_rate": 0.00015,
+      "loss": 4.658082008361816,
+      "step": 880
+    },
+    {
+      "epoch": 0.2678015046736074,
+      "grad_norm": 1.714073896408081,
+      "learning_rate": 0.00015,
+      "loss": 4.648862361907959,
+      "step": 881
+    },
+    {
+      "epoch": 0.26810547913975225,
+      "grad_norm": 1.9790101051330566,
+      "learning_rate": 0.00015,
+      "loss": 4.483827590942383,
+      "step": 882
+    },
+    {
+      "epoch": 0.26840945360589713,
+      "grad_norm": 1.8374898433685303,
+      "learning_rate": 0.00015,
+      "loss": 4.289336204528809,
+      "step": 883
+    },
+    {
+      "epoch": 0.26871342807204196,
+      "grad_norm": 1.8189436197280884,
+      "learning_rate": 0.00015,
+      "loss": 4.751187324523926,
+      "step": 884
+    },
+    {
+      "epoch": 0.2690174025381868,
+      "grad_norm": 2.555690050125122,
+      "learning_rate": 0.00015,
+      "loss": 4.864801406860352,
+      "step": 885
+    },
+    {
+      "epoch": 0.2693213770043316,
+      "grad_norm": 2.3488893508911133,
+      "learning_rate": 0.00015,
+      "loss": 4.651714324951172,
+      "step": 886
+    },
+    {
+      "epoch": 0.2696253514704765,
+      "grad_norm": 2.041745662689209,
+      "learning_rate": 0.00015,
+      "loss": 4.304879188537598,
+      "step": 887
+    },
+    {
+      "epoch": 0.26992932593662133,
+      "grad_norm": 1.7490661144256592,
+      "learning_rate": 0.00015,
+      "loss": 4.5656843185424805,
+      "step": 888
+    },
+    {
+      "epoch": 0.27023330040276616,
+      "grad_norm": 2.0361738204956055,
+      "learning_rate": 0.00015,
+      "loss": 4.607753753662109,
+      "step": 889
+    },
+    {
+      "epoch": 0.270537274868911,
+      "grad_norm": 2.050398826599121,
+      "learning_rate": 0.00015,
+      "loss": 4.265903472900391,
+      "step": 890
+    },
+    {
+      "epoch": 0.2708412493350559,
+      "grad_norm": 2.373138427734375,
+      "learning_rate": 0.00015,
+      "loss": 4.420685768127441,
+      "step": 891
+    },
+    {
+      "epoch": 0.2711452238012007,
+      "grad_norm": 1.749476671218872,
+      "learning_rate": 0.00015,
+      "loss": 4.532936096191406,
+      "step": 892
+    },
+    {
+      "epoch": 0.27144919826734554,
+      "grad_norm": 2.086395502090454,
+      "learning_rate": 0.00015,
+      "loss": 4.625255584716797,
+      "step": 893
+    },
+    {
+      "epoch": 0.27175317273349037,
+      "grad_norm": 2.0593135356903076,
+      "learning_rate": 0.00015,
+      "loss": 4.420129776000977,
+      "step": 894
+    },
+    {
+      "epoch": 0.27205714719963525,
+      "grad_norm": 2.266852378845215,
+      "learning_rate": 0.00015,
+      "loss": 4.408595561981201,
+      "step": 895
+    },
+    {
+      "epoch": 0.2723611216657801,
+      "grad_norm": 2.2006614208221436,
+      "learning_rate": 0.00015,
+      "loss": 4.373136043548584,
+      "step": 896
+    },
+    {
+      "epoch": 0.2726650961319249,
+      "grad_norm": 3.2129135131835938,
+      "learning_rate": 0.00015,
+      "loss": 4.549173831939697,
+      "step": 897
+    },
+    {
+      "epoch": 0.27296907059806974,
+      "grad_norm": 2.178008794784546,
+      "learning_rate": 0.00015,
+      "loss": 4.296047687530518,
+      "step": 898
+    },
+    {
+      "epoch": 0.2732730450642146,
+      "grad_norm": 1.921857237815857,
+      "learning_rate": 0.00015,
+      "loss": 4.542671203613281,
+      "step": 899
+    },
+    {
+      "epoch": 0.27357701953035946,
+      "grad_norm": 2.1875381469726562,
+      "learning_rate": 0.00015,
+      "loss": 4.600428581237793,
+      "step": 900
+    },
+    {
+      "epoch": 0.2738809939965043,
+      "grad_norm": 2.535599708557129,
+      "learning_rate": 0.00015,
+      "loss": 4.199663162231445,
+      "step": 901
+    },
+    {
+      "epoch": 0.2741849684626491,
+      "grad_norm": 2.2965691089630127,
+      "learning_rate": 0.00015,
+      "loss": 4.564335346221924,
+      "step": 902
+    },
+    {
+      "epoch": 0.274488942928794,
+      "grad_norm": 1.6621229648590088,
+      "learning_rate": 0.00015,
+      "loss": 4.403615951538086,
+      "step": 903
+    },
+    {
+      "epoch": 0.27479291739493883,
+      "grad_norm": 2.0944924354553223,
+      "learning_rate": 0.00015,
+      "loss": 4.768467426300049,
+      "step": 904
+    },
+    {
+      "epoch": 0.27509689186108366,
+      "grad_norm": 2.654827356338501,
+      "learning_rate": 0.00015,
+      "loss": 4.737344741821289,
+      "step": 905
+    },
+    {
+      "epoch": 0.2754008663272285,
+      "grad_norm": 2.053867816925049,
+      "learning_rate": 0.00015,
+      "loss": 4.115928649902344,
+      "step": 906
+    },
+    {
+      "epoch": 0.2757048407933734,
+      "grad_norm": 2.295719623565674,
+      "learning_rate": 0.00015,
+      "loss": 4.447504043579102,
+      "step": 907
+    },
+    {
+      "epoch": 0.2760088152595182,
+      "grad_norm": 2.496553659439087,
+      "learning_rate": 0.00015,
+      "loss": 4.224276542663574,
+      "step": 908
+    },
+    {
+      "epoch": 0.27631278972566303,
+      "grad_norm": 1.9621105194091797,
+      "learning_rate": 0.00015,
+      "loss": 4.245770454406738,
+      "step": 909
+    },
+    {
+      "epoch": 0.27661676419180786,
+      "grad_norm": 2.781357765197754,
+      "learning_rate": 0.00015,
+      "loss": 4.5496368408203125,
+      "step": 910
+    },
+    {
+      "epoch": 0.27692073865795275,
+      "grad_norm": 2.350013017654419,
+      "learning_rate": 0.00015,
+      "loss": 4.493282794952393,
+      "step": 911
+    },
+    {
+      "epoch": 0.2772247131240976,
+      "grad_norm": 1.8348884582519531,
+      "learning_rate": 0.00015,
+      "loss": 4.557445049285889,
+      "step": 912
+    },
+    {
+      "epoch": 0.2775286875902424,
+      "grad_norm": 2.001253366470337,
+      "learning_rate": 0.00015,
+      "loss": 4.949978828430176,
+      "step": 913
+    },
+    {
+      "epoch": 0.27783266205638724,
+      "grad_norm": 1.8087342977523804,
+      "learning_rate": 0.00015,
+      "loss": 4.347936630249023,
+      "step": 914
+    },
+    {
+      "epoch": 0.2781366365225321,
+      "grad_norm": 2.3734161853790283,
+      "learning_rate": 0.00015,
+      "loss": 4.602643966674805,
+      "step": 915
+    },
+    {
+      "epoch": 0.27844061098867695,
+      "grad_norm": 2.4027411937713623,
+      "learning_rate": 0.00015,
+      "loss": 4.486043930053711,
+      "step": 916
+    },
+    {
+      "epoch": 0.2787445854548218,
+      "grad_norm": 2.735114574432373,
+      "learning_rate": 0.00015,
+      "loss": 4.659709930419922,
+      "step": 917
+    },
+    {
+      "epoch": 0.27904855992096667,
+      "grad_norm": 2.7607638835906982,
+      "learning_rate": 0.00015,
+      "loss": 4.243941307067871,
+      "step": 918
+    },
+    {
+      "epoch": 0.2793525343871115,
+      "grad_norm": 1.9252501726150513,
+      "learning_rate": 0.00015,
+      "loss": 4.245811462402344,
+      "step": 919
+    },
+    {
+      "epoch": 0.2796565088532563,
+      "grad_norm": 2.0722951889038086,
+      "learning_rate": 0.00015,
+      "loss": 4.361391067504883,
+      "step": 920
+    },
+    {
+      "epoch": 0.27996048331940115,
+      "grad_norm": 2.2559621334075928,
+      "learning_rate": 0.00015,
+      "loss": 4.256713390350342,
+      "step": 921
+    },
+    {
+      "epoch": 0.28026445778554604,
+      "grad_norm": 1.8118853569030762,
+      "learning_rate": 0.00015,
+      "loss": 4.493771553039551,
+      "step": 922
+    },
+    {
+      "epoch": 0.28056843225169087,
+      "grad_norm": 1.745952844619751,
+      "learning_rate": 0.00015,
+      "loss": 4.477970123291016,
+      "step": 923
+    },
+    {
+      "epoch": 0.2808724067178357,
+      "grad_norm": 1.8047990798950195,
+      "learning_rate": 0.00015,
+      "loss": 4.415807247161865,
+      "step": 924
+    },
+    {
+      "epoch": 0.28117638118398053,
+      "grad_norm": 1.8159899711608887,
+      "learning_rate": 0.00015,
+      "loss": 4.534528732299805,
+      "step": 925
+    },
+    {
+      "epoch": 0.2814803556501254,
+      "grad_norm": 3.2453927993774414,
+      "learning_rate": 0.00015,
+      "loss": 4.386707305908203,
+      "step": 926
+    },
+    {
+      "epoch": 0.28178433011627024,
+      "grad_norm": 1.8058236837387085,
+      "learning_rate": 0.00015,
+      "loss": 4.4044389724731445,
+      "step": 927
+    },
+    {
+      "epoch": 0.2820883045824151,
+      "grad_norm": 1.890562653541565,
+      "learning_rate": 0.00015,
+      "loss": 4.302713394165039,
+      "step": 928
+    },
+    {
+      "epoch": 0.2823922790485599,
+      "grad_norm": 2.780857563018799,
+      "learning_rate": 0.00015,
+      "loss": 4.398674488067627,
+      "step": 929
+    },
+    {
+      "epoch": 0.2826962535147048,
+      "grad_norm": 4.453455448150635,
+      "learning_rate": 0.00015,
+      "loss": 4.684018611907959,
+      "step": 930
+    },
+    {
+      "epoch": 0.2830002279808496,
+      "grad_norm": 1.9640278816223145,
+      "learning_rate": 0.00015,
+      "loss": 4.580140590667725,
+      "step": 931
+    },
+    {
+      "epoch": 0.28330420244699445,
+      "grad_norm": 1.7051184177398682,
+      "learning_rate": 0.00015,
+      "loss": 4.596173286437988,
+      "step": 932
+    },
+    {
+      "epoch": 0.2836081769131393,
+      "grad_norm": 1.6775091886520386,
+      "learning_rate": 0.00015,
+      "loss": 4.487752437591553,
+      "step": 933
+    },
+    {
+      "epoch": 0.28391215137928416,
+      "grad_norm": 2.4020729064941406,
+      "learning_rate": 0.00015,
+      "loss": 4.539383411407471,
+      "step": 934
+    },
+    {
+      "epoch": 0.284216125845429,
+      "grad_norm": 1.8002465963363647,
+      "learning_rate": 0.00015,
+      "loss": 4.467827796936035,
+      "step": 935
+    },
+    {
+      "epoch": 0.2845201003115738,
+      "grad_norm": 1.8578530550003052,
+      "learning_rate": 0.00015,
+      "loss": 4.374014854431152,
+      "step": 936
+    },
+    {
+      "epoch": 0.28482407477771865,
+      "grad_norm": 2.1210553646087646,
+      "learning_rate": 0.00015,
+      "loss": 4.552811622619629,
+      "step": 937
+    },
+    {
+      "epoch": 0.28512804924386354,
+      "grad_norm": 1.731631875038147,
+      "learning_rate": 0.00015,
+      "loss": 4.328273296356201,
+      "step": 938
+    },
+    {
+      "epoch": 0.28543202371000836,
+      "grad_norm": 2.8373637199401855,
+      "learning_rate": 0.00015,
+      "loss": 4.754038333892822,
+      "step": 939
+    },
+    {
+      "epoch": 0.2857359981761532,
+      "grad_norm": 1.8735849857330322,
+      "learning_rate": 0.00015,
+      "loss": 4.4415130615234375,
+      "step": 940
+    },
+    {
+      "epoch": 0.286039972642298,
+      "grad_norm": 2.0340423583984375,
+      "learning_rate": 0.00015,
+      "loss": 4.775668144226074,
+      "step": 941
+    },
+    {
+      "epoch": 0.2863439471084429,
+      "grad_norm": 1.8985058069229126,
+      "learning_rate": 0.00015,
+      "loss": 4.3031816482543945,
+      "step": 942
+    },
+    {
+      "epoch": 0.28664792157458774,
+      "grad_norm": 1.893847107887268,
+      "learning_rate": 0.00015,
+      "loss": 4.4812774658203125,
+      "step": 943
+    },
+    {
+      "epoch": 0.28695189604073257,
+      "grad_norm": 1.992106556892395,
+      "learning_rate": 0.00015,
+      "loss": 4.48546028137207,
+      "step": 944
+    },
+    {
+      "epoch": 0.2872558705068774,
+      "grad_norm": 2.1293983459472656,
+      "learning_rate": 0.00015,
+      "loss": 4.423661708831787,
+      "step": 945
+    },
+    {
+      "epoch": 0.2875598449730223,
+      "grad_norm": 2.3705222606658936,
+      "learning_rate": 0.00015,
+      "loss": 4.652749538421631,
+      "step": 946
+    },
+    {
+      "epoch": 0.2878638194391671,
+      "grad_norm": 2.1059329509735107,
+      "learning_rate": 0.00015,
+      "loss": 4.455974102020264,
+      "step": 947
+    },
+    {
+      "epoch": 0.28816779390531194,
+      "grad_norm": 2.078476905822754,
+      "learning_rate": 0.00015,
+      "loss": 4.254177093505859,
+      "step": 948
+    },
+    {
+      "epoch": 0.28847176837145677,
+      "grad_norm": 1.981514573097229,
+      "learning_rate": 0.00015,
+      "loss": 4.601126670837402,
+      "step": 949
+    },
+    {
+      "epoch": 0.28877574283760166,
+      "grad_norm": 2.7646474838256836,
+      "learning_rate": 0.00015,
+      "loss": 4.343170166015625,
+      "step": 950
+    },
+    {
+      "epoch": 0.2890797173037465,
+      "grad_norm": 2.0774126052856445,
+      "learning_rate": 0.00015,
+      "loss": 4.310368537902832,
+      "step": 951
+    },
+    {
+      "epoch": 0.2893836917698913,
+      "grad_norm": 2.2458932399749756,
+      "learning_rate": 0.00015,
+      "loss": 4.295446395874023,
+      "step": 952
+    },
+    {
+      "epoch": 0.28968766623603615,
+      "grad_norm": 2.0657401084899902,
+      "learning_rate": 0.00015,
+      "loss": 4.366024971008301,
+      "step": 953
+    },
+    {
+      "epoch": 0.28999164070218103,
+      "grad_norm": 1.7287461757659912,
+      "learning_rate": 0.00015,
+      "loss": 4.5319108963012695,
+      "step": 954
+    },
+    {
+      "epoch": 0.29029561516832586,
+      "grad_norm": 2.3741533756256104,
+      "learning_rate": 0.00015,
+      "loss": 4.702271461486816,
+      "step": 955
+    },
+    {
+      "epoch": 0.2905995896344707,
+      "grad_norm": 1.857850432395935,
+      "learning_rate": 0.00015,
+      "loss": 4.219401836395264,
+      "step": 956
+    },
+    {
+      "epoch": 0.2909035641006156,
+      "grad_norm": 1.978426218032837,
+      "learning_rate": 0.00015,
+      "loss": 4.321810722351074,
+      "step": 957
+    },
+    {
+      "epoch": 0.2912075385667604,
+      "grad_norm": 2.1086721420288086,
+      "learning_rate": 0.00015,
+      "loss": 4.328747272491455,
+      "step": 958
+    },
+    {
+      "epoch": 0.29151151303290523,
+      "grad_norm": 2.2524843215942383,
+      "learning_rate": 0.00015,
+      "loss": 4.518558979034424,
+      "step": 959
+    },
+    {
+      "epoch": 0.29181548749905006,
+      "grad_norm": 2.0697712898254395,
+      "learning_rate": 0.00015,
+      "loss": 4.384339332580566,
+      "step": 960
+    },
+    {
+      "epoch": 0.29211946196519495,
+      "grad_norm": 2.0855631828308105,
+      "learning_rate": 0.00015,
+      "loss": 4.618060111999512,
+      "step": 961
+    },
+    {
+      "epoch": 0.2924234364313398,
+      "grad_norm": 1.969549536705017,
+      "learning_rate": 0.00015,
+      "loss": 4.121874809265137,
+      "step": 962
+    },
+    {
+      "epoch": 0.2927274108974846,
+      "grad_norm": 1.7161803245544434,
+      "learning_rate": 0.00015,
+      "loss": 4.426275253295898,
+      "step": 963
+    },
+    {
+      "epoch": 0.29303138536362944,
+      "grad_norm": 1.8243383169174194,
+      "learning_rate": 0.00015,
+      "loss": 4.342365264892578,
+      "step": 964
+    },
+    {
+      "epoch": 0.2933353598297743,
+      "grad_norm": 1.9362605810165405,
+      "learning_rate": 0.00015,
+      "loss": 4.329897880554199,
+      "step": 965
+    },
+    {
+      "epoch": 0.29363933429591915,
+      "grad_norm": 4.13656759262085,
+      "learning_rate": 0.00015,
+      "loss": 5.019277572631836,
+      "step": 966
+    },
+    {
+      "epoch": 0.293943308762064,
+      "grad_norm": 2.3126707077026367,
+      "learning_rate": 0.00015,
+      "loss": 4.576555252075195,
+      "step": 967
+    },
+    {
+      "epoch": 0.2942472832282088,
+      "grad_norm": 2.0928938388824463,
+      "learning_rate": 0.00015,
+      "loss": 4.603344440460205,
+      "step": 968
+    },
+    {
+      "epoch": 0.2945512576943537,
+      "grad_norm": 2.236098289489746,
+      "learning_rate": 0.00015,
+      "loss": 4.289191246032715,
+      "step": 969
+    },
+    {
+      "epoch": 0.2948552321604985,
+      "grad_norm": 2.2182371616363525,
+      "learning_rate": 0.00015,
+      "loss": 4.257903099060059,
+      "step": 970
+    },
+    {
+      "epoch": 0.29515920662664336,
+      "grad_norm": 1.9866735935211182,
+      "learning_rate": 0.00015,
+      "loss": 4.535268783569336,
+      "step": 971
+    },
+    {
+      "epoch": 0.2954631810927882,
+      "grad_norm": 1.8852744102478027,
+      "learning_rate": 0.00015,
+      "loss": 4.788799285888672,
+      "step": 972
+    },
+    {
+      "epoch": 0.29576715555893307,
+      "grad_norm": 2.092540979385376,
+      "learning_rate": 0.00015,
+      "loss": 4.583626747131348,
+      "step": 973
+    },
+    {
+      "epoch": 0.2960711300250779,
+      "grad_norm": 2.283473253250122,
+      "learning_rate": 0.00015,
+      "loss": 4.434234142303467,
+      "step": 974
+    },
+    {
+      "epoch": 0.29637510449122273,
+      "grad_norm": 1.764085292816162,
+      "learning_rate": 0.00015,
+      "loss": 4.7144975662231445,
+      "step": 975
+    },
+    {
+      "epoch": 0.29667907895736756,
+      "grad_norm": 1.8126758337020874,
+      "learning_rate": 0.00015,
+      "loss": 4.376204490661621,
+      "step": 976
+    },
+    {
+      "epoch": 0.29698305342351244,
+      "grad_norm": 1.9786795377731323,
+      "learning_rate": 0.00015,
+      "loss": 4.495676040649414,
+      "step": 977
+    },
+    {
+      "epoch": 0.2972870278896573,
+      "grad_norm": 2.133612632751465,
+      "learning_rate": 0.00015,
+      "loss": 4.497121810913086,
+      "step": 978
+    },
+    {
+      "epoch": 0.2975910023558021,
+      "grad_norm": 2.107250213623047,
+      "learning_rate": 0.00015,
+      "loss": 4.326852798461914,
+      "step": 979
+    },
+    {
+      "epoch": 0.29789497682194693,
+      "grad_norm": 1.9591680765151978,
+      "learning_rate": 0.00015,
+      "loss": 4.2498393058776855,
+      "step": 980
+    },
+    {
+      "epoch": 0.2981989512880918,
+      "grad_norm": 2.2823190689086914,
+      "learning_rate": 0.00015,
+      "loss": 4.407077789306641,
+      "step": 981
+    },
+    {
+      "epoch": 0.29850292575423665,
+      "grad_norm": 2.323568105697632,
+      "learning_rate": 0.00015,
+      "loss": 4.705851078033447,
+      "step": 982
+    },
+    {
+      "epoch": 0.2988069002203815,
+      "grad_norm": 1.9538623094558716,
+      "learning_rate": 0.00015,
+      "loss": 4.26961612701416,
+      "step": 983
+    },
+    {
+      "epoch": 0.2991108746865263,
+      "grad_norm": 2.2118146419525146,
+      "learning_rate": 0.00015,
+      "loss": 4.354861736297607,
+      "step": 984
+    },
+    {
+      "epoch": 0.2994148491526712,
+      "grad_norm": 2.0708906650543213,
+      "learning_rate": 0.00015,
+      "loss": 4.234423637390137,
+      "step": 985
+    },
+    {
+      "epoch": 0.299718823618816,
+      "grad_norm": 1.9159834384918213,
+      "learning_rate": 0.00015,
+      "loss": 4.021456241607666,
+      "step": 986
+    },
+    {
+      "epoch": 0.30002279808496085,
+      "grad_norm": 1.9587281942367554,
+      "learning_rate": 0.00015,
+      "loss": 4.664050579071045,
+      "step": 987
+    },
+    {
+      "epoch": 0.3003267725511057,
+      "grad_norm": 1.995664358139038,
+      "learning_rate": 0.00015,
+      "loss": 4.547416687011719,
+      "step": 988
+    },
+    {
+      "epoch": 0.30063074701725057,
+      "grad_norm": 2.0282983779907227,
+      "learning_rate": 0.00015,
+      "loss": 4.526060581207275,
+      "step": 989
+    },
+    {
+      "epoch": 0.3009347214833954,
+      "grad_norm": 1.8454768657684326,
+      "learning_rate": 0.00015,
+      "loss": 4.431363105773926,
+      "step": 990
+    },
+    {
+      "epoch": 0.3012386959495402,
+      "grad_norm": 1.8858832120895386,
+      "learning_rate": 0.00015,
+      "loss": 4.271605014801025,
+      "step": 991
+    },
+    {
+      "epoch": 0.3015426704156851,
+      "grad_norm": 1.9861679077148438,
+      "learning_rate": 0.00015,
+      "loss": 4.515031814575195,
+      "step": 992
+    },
+    {
+      "epoch": 0.30184664488182994,
+      "grad_norm": 1.9714046716690063,
+      "learning_rate": 0.00015,
+      "loss": 4.699455738067627,
+      "step": 993
+    },
+    {
+      "epoch": 0.30215061934797477,
+      "grad_norm": 4.38908052444458,
+      "learning_rate": 0.00015,
+      "loss": 4.612029552459717,
+      "step": 994
+    },
+    {
+      "epoch": 0.3024545938141196,
+      "grad_norm": 2.1126561164855957,
+      "learning_rate": 0.00015,
+      "loss": 4.674310684204102,
+      "step": 995
+    },
+    {
+      "epoch": 0.3027585682802645,
+      "grad_norm": 1.922357439994812,
+      "learning_rate": 0.00015,
+      "loss": 4.389533042907715,
+      "step": 996
+    },
+    {
+      "epoch": 0.3030625427464093,
+      "grad_norm": 2.242847204208374,
+      "learning_rate": 0.00015,
+      "loss": 4.2144880294799805,
+      "step": 997
+    },
+    {
+      "epoch": 0.30336651721255414,
+      "grad_norm": 2.666672468185425,
+      "learning_rate": 0.00015,
+      "loss": 4.386487007141113,
+      "step": 998
+    },
+    {
+      "epoch": 0.303670491678699,
+      "grad_norm": 1.8454669713974,
+      "learning_rate": 0.00015,
+      "loss": 4.5814666748046875,
+      "step": 999
+    },
+    {
+      "epoch": 0.30397446614484386,
+      "grad_norm": 2.3694045543670654,
+      "learning_rate": 0.00015,
+      "loss": 4.178978443145752,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3042784406109887,
+      "grad_norm": 2.312467098236084,
+      "learning_rate": 0.00015,
+      "loss": 4.422657012939453,
+      "step": 1001
+    },
+    {
+      "epoch": 0.3045824150771335,
+      "grad_norm": 3.7126100063323975,
+      "learning_rate": 0.00015,
+      "loss": 4.5202226638793945,
+      "step": 1002
+    },
+    {
+      "epoch": 0.30488638954327835,
+      "grad_norm": 2.2542130947113037,
+      "learning_rate": 0.00015,
+      "loss": 4.388377666473389,
+      "step": 1003
+    },
+    {
+      "epoch": 0.30519036400942323,
+      "grad_norm": 1.849789023399353,
+      "learning_rate": 0.00015,
+      "loss": 4.484010696411133,
+      "step": 1004
+    },
+    {
+      "epoch": 0.30549433847556806,
+      "grad_norm": 2.5572357177734375,
+      "learning_rate": 0.00015,
+      "loss": 4.363813400268555,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3057983129417129,
+      "grad_norm": 2.3575239181518555,
+      "learning_rate": 0.00015,
+      "loss": 4.292799472808838,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3061022874078577,
+      "grad_norm": 2.073882818222046,
+      "learning_rate": 0.00015,
+      "loss": 4.408339500427246,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3064062618740026,
+      "grad_norm": 1.9969879388809204,
+      "learning_rate": 0.00015,
+      "loss": 4.346457481384277,
+      "step": 1008
+    },
+    {
+      "epoch": 0.30671023634014744,
+      "grad_norm": 2.208486318588257,
+      "learning_rate": 0.00015,
+      "loss": 4.36622428894043,
+      "step": 1009
+    },
+    {
+      "epoch": 0.30701421080629226,
+      "grad_norm": 2.230905294418335,
+      "learning_rate": 0.00015,
+      "loss": 4.14420747756958,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3073181852724371,
+      "grad_norm": 2.5348477363586426,
+      "learning_rate": 0.00015,
+      "loss": 4.41834831237793,
+      "step": 1011
+    },
+    {
+      "epoch": 0.307622159738582,
+      "grad_norm": 2.2350683212280273,
+      "learning_rate": 0.00015,
+      "loss": 4.381241798400879,
+      "step": 1012
+    },
+    {
+      "epoch": 0.3079261342047268,
+      "grad_norm": 2.1959993839263916,
+      "learning_rate": 0.00015,
+      "loss": 4.284981727600098,
+      "step": 1013
+    },
+    {
+      "epoch": 0.30823010867087164,
+      "grad_norm": 1.982353925704956,
+      "learning_rate": 0.00015,
+      "loss": 4.431988716125488,
+      "step": 1014
+    },
+    {
+      "epoch": 0.30853408313701647,
+      "grad_norm": 2.0439248085021973,
+      "learning_rate": 0.00015,
+      "loss": 4.456663131713867,
+      "step": 1015
+    },
+    {
+      "epoch": 0.30883805760316135,
+      "grad_norm": 2.0135128498077393,
+      "learning_rate": 0.00015,
+      "loss": 4.260210037231445,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3091420320693062,
+      "grad_norm": 1.5713000297546387,
+      "learning_rate": 0.00015,
+      "loss": 4.405777931213379,
+      "step": 1017
+    },
+    {
+      "epoch": 0.309446006535451,
+      "grad_norm": 2.1084401607513428,
+      "learning_rate": 0.00015,
+      "loss": 4.463507652282715,
+      "step": 1018
+    },
+    {
+      "epoch": 0.30974998100159584,
+      "grad_norm": 1.916013240814209,
+      "learning_rate": 0.00015,
+      "loss": 4.340442180633545,
+      "step": 1019
+    },
+    {
+      "epoch": 0.3100539554677407,
+      "grad_norm": 2.2796170711517334,
+      "learning_rate": 0.00015,
+      "loss": 4.280055046081543,
+      "step": 1020
+    },
+    {
+      "epoch": 0.31035792993388556,
+      "grad_norm": 2.458641767501831,
+      "learning_rate": 0.00015,
+      "loss": 4.127218246459961,
+      "step": 1021
+    },
+    {
+      "epoch": 0.3106619044000304,
+      "grad_norm": 2.191302537918091,
+      "learning_rate": 0.00015,
+      "loss": 4.840422630310059,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3109658788661752,
+      "grad_norm": 2.2489781379699707,
+      "learning_rate": 0.00015,
+      "loss": 4.711808204650879,
+      "step": 1023
+    },
+    {
+      "epoch": 0.3112698533323201,
+      "grad_norm": 1.7385538816452026,
+      "learning_rate": 0.00015,
+      "loss": 4.307732582092285,
+      "step": 1024
+    },
+    {
+      "epoch": 0.31157382779846493,
+      "grad_norm": 2.319798707962036,
+      "learning_rate": 0.00015,
+      "loss": 4.322588920593262,
+      "step": 1025
+    },
+    {
+      "epoch": 0.31187780226460976,
+      "grad_norm": 2.0708072185516357,
+      "learning_rate": 0.00015,
+      "loss": 4.16618013381958,
+      "step": 1026
+    },
+    {
+      "epoch": 0.3121817767307546,
+      "grad_norm": 2.2142839431762695,
+      "learning_rate": 0.00015,
+      "loss": 4.3772478103637695,
+      "step": 1027
+    },
+    {
+      "epoch": 0.3124857511968995,
+      "grad_norm": 1.7371846437454224,
+      "learning_rate": 0.00015,
+      "loss": 4.430246829986572,
+      "step": 1028
+    },
+    {
+      "epoch": 0.3127897256630443,
+      "grad_norm": 2.0236129760742188,
+      "learning_rate": 0.00015,
+      "loss": 4.569117546081543,
+      "step": 1029
+    },
+    {
+      "epoch": 0.31309370012918913,
+      "grad_norm": 2.308443546295166,
+      "learning_rate": 0.00015,
+      "loss": 4.903112411499023,
+      "step": 1030
+    },
+    {
+      "epoch": 0.313397674595334,
+      "grad_norm": 2.16902232170105,
+      "learning_rate": 0.00015,
+      "loss": 4.742799758911133,
+      "step": 1031
+    },
+    {
+      "epoch": 0.31370164906147885,
+      "grad_norm": 2.1175599098205566,
+      "learning_rate": 0.00015,
+      "loss": 4.729710102081299,
+      "step": 1032
+    },
+    {
+      "epoch": 0.3140056235276237,
+      "grad_norm": 2.12835693359375,
+      "learning_rate": 0.00015,
+      "loss": 4.312872409820557,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3143095979937685,
+      "grad_norm": 2.212897300720215,
+      "learning_rate": 0.00015,
+      "loss": 4.302469253540039,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3146135724599134,
+      "grad_norm": 2.322291135787964,
+      "learning_rate": 0.00015,
+      "loss": 4.3713579177856445,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3149175469260582,
+      "grad_norm": 2.0713953971862793,
+      "learning_rate": 0.00015,
+      "loss": 4.376710414886475,
+      "step": 1036
+    },
+    {
+      "epoch": 0.31522152139220305,
+      "grad_norm": 2.122255325317383,
+      "learning_rate": 0.00015,
+      "loss": 4.490863800048828,
+      "step": 1037
+    },
+    {
+      "epoch": 0.3155254958583479,
+      "grad_norm": 1.8258548974990845,
+      "learning_rate": 0.00015,
+      "loss": 4.268899917602539,
+      "step": 1038
+    },
+    {
+      "epoch": 0.31582947032449277,
+      "grad_norm": 2.1500778198242188,
+      "learning_rate": 0.00015,
+      "loss": 4.50262451171875,
+      "step": 1039
+    },
+    {
+      "epoch": 0.3161334447906376,
+      "grad_norm": 2.7903459072113037,
+      "learning_rate": 0.00015,
+      "loss": 4.501411437988281,
+      "step": 1040
+    },
+    {
+      "epoch": 0.3164374192567824,
+      "grad_norm": 2.0578207969665527,
+      "learning_rate": 0.00015,
+      "loss": 4.40083646774292,
+      "step": 1041
+    },
+    {
+      "epoch": 0.31674139372292726,
+      "grad_norm": 2.0132863521575928,
+      "learning_rate": 0.00015,
+      "loss": 4.479395389556885,
+      "step": 1042
+    },
+    {
+      "epoch": 0.31704536818907214,
+      "grad_norm": 1.868908166885376,
+      "learning_rate": 0.00015,
+      "loss": 4.432626724243164,
+      "step": 1043
+    },
+    {
+      "epoch": 0.31734934265521697,
+      "grad_norm": 2.0285027027130127,
+      "learning_rate": 0.00015,
+      "loss": 4.471344947814941,
+      "step": 1044
+    },
+    {
+      "epoch": 0.3176533171213618,
+      "grad_norm": 2.0714588165283203,
+      "learning_rate": 0.00015,
+      "loss": 4.327896595001221,
+      "step": 1045
+    },
+    {
+      "epoch": 0.31795729158750663,
+      "grad_norm": 1.859429121017456,
+      "learning_rate": 0.00015,
+      "loss": 4.5692362785339355,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3182612660536515,
+      "grad_norm": 1.7299549579620361,
+      "learning_rate": 0.00015,
+      "loss": 4.6055474281311035,
+      "step": 1047
+    },
+    {
+      "epoch": 0.31856524051979634,
+      "grad_norm": 2.0733907222747803,
+      "learning_rate": 0.00015,
+      "loss": 4.083170413970947,
+      "step": 1048
+    },
+    {
+      "epoch": 0.3188692149859412,
+      "grad_norm": 2.071382522583008,
+      "learning_rate": 0.00015,
+      "loss": 4.489770889282227,
+      "step": 1049
+    },
+    {
+      "epoch": 0.319173189452086,
+      "grad_norm": 2.4808435440063477,
+      "learning_rate": 0.00015,
+      "loss": 4.4711198806762695,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3194771639182309,
+      "grad_norm": 1.9795262813568115,
+      "learning_rate": 0.00015,
+      "loss": 4.308687210083008,
+      "step": 1051
+    },
+    {
+      "epoch": 0.3197811383843757,
+      "grad_norm": 1.5751978158950806,
+      "learning_rate": 0.00015,
+      "loss": 4.4411940574646,
+      "step": 1052
+    },
+    {
+      "epoch": 0.32008511285052055,
+      "grad_norm": 1.9910207986831665,
+      "learning_rate": 0.00015,
+      "loss": 4.723758697509766,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3203890873166654,
+      "grad_norm": 1.6922791004180908,
+      "learning_rate": 0.00015,
+      "loss": 4.312475204467773,
+      "step": 1054
+    },
+    {
+      "epoch": 0.32069306178281026,
+      "grad_norm": 1.801006555557251,
+      "learning_rate": 0.00015,
+      "loss": 4.140463352203369,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3209970362489551,
+      "grad_norm": 1.7544379234313965,
+      "learning_rate": 0.00015,
+      "loss": 4.463098526000977,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3213010107150999,
+      "grad_norm": 2.3533692359924316,
+      "learning_rate": 0.00015,
+      "loss": 4.337651252746582,
+      "step": 1057
+    },
+    {
+      "epoch": 0.32160498518124475,
+      "grad_norm": 1.7905725240707397,
+      "learning_rate": 0.00015,
+      "loss": 4.109731674194336,
+      "step": 1058
+    },
+    {
+      "epoch": 0.32190895964738964,
+      "grad_norm": 1.9509469270706177,
+      "learning_rate": 0.00015,
+      "loss": 4.487770080566406,
+      "step": 1059
+    },
+    {
+      "epoch": 0.32221293411353447,
+      "grad_norm": 2.0259923934936523,
+      "learning_rate": 0.00015,
+      "loss": 4.4123735427856445,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3225169085796793,
+      "grad_norm": 2.0444247722625732,
+      "learning_rate": 0.00015,
+      "loss": 4.20911979675293,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3228208830458241,
+      "grad_norm": 1.77584707736969,
+      "learning_rate": 0.00015,
+      "loss": 4.289585590362549,
+      "step": 1062
+    },
+    {
+      "epoch": 0.323124857511969,
+      "grad_norm": 2.0790975093841553,
+      "learning_rate": 0.00015,
+      "loss": 4.433958530426025,
+      "step": 1063
+    },
+    {
+      "epoch": 0.32342883197811384,
+      "grad_norm": 2.296135187149048,
+      "learning_rate": 0.00015,
+      "loss": 4.344565391540527,
+      "step": 1064
+    },
+    {
+      "epoch": 0.32373280644425867,
+      "grad_norm": 2.3378243446350098,
+      "learning_rate": 0.00015,
+      "loss": 4.397375106811523,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3240367809104035,
+      "grad_norm": 1.8747236728668213,
+      "learning_rate": 0.00015,
+      "loss": 4.390880584716797,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3243407553765484,
+      "grad_norm": 1.987847089767456,
+      "learning_rate": 0.00015,
+      "loss": 4.456721305847168,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3246447298426932,
+      "grad_norm": 2.0660431385040283,
+      "learning_rate": 0.00015,
+      "loss": 4.1006317138671875,
+      "step": 1068
+    },
+    {
+      "epoch": 0.32494870430883804,
+      "grad_norm": 3.064868927001953,
+      "learning_rate": 0.00015,
+      "loss": 4.706048011779785,
+      "step": 1069
+    },
+    {
+      "epoch": 0.32525267877498293,
+      "grad_norm": 1.9179483652114868,
+      "learning_rate": 0.00015,
+      "loss": 4.612796783447266,
+      "step": 1070
+    },
+    {
+      "epoch": 0.32555665324112776,
+      "grad_norm": 1.6911954879760742,
+      "learning_rate": 0.00015,
+      "loss": 4.202427864074707,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3258606277072726,
+      "grad_norm": 2.049736738204956,
+      "learning_rate": 0.00015,
+      "loss": 4.093686580657959,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3261646021734174,
+      "grad_norm": 1.7808513641357422,
+      "learning_rate": 0.00015,
+      "loss": 4.213599681854248,
+      "step": 1073
+    },
+    {
+      "epoch": 0.3264685766395623,
+      "grad_norm": 2.7623167037963867,
+      "learning_rate": 0.00015,
+      "loss": 4.3716254234313965,
+      "step": 1074
+    },
+    {
+      "epoch": 0.32677255110570713,
+      "grad_norm": 2.047752618789673,
+      "learning_rate": 0.00015,
+      "loss": 4.397392272949219,
+      "step": 1075
+    },
+    {
+      "epoch": 0.32707652557185196,
+      "grad_norm": 1.9015398025512695,
+      "learning_rate": 0.00015,
+      "loss": 4.345738887786865,
+      "step": 1076
+    },
+    {
+      "epoch": 0.3273805000379968,
+      "grad_norm": 1.6867552995681763,
+      "learning_rate": 0.00015,
+      "loss": 4.4111127853393555,
+      "step": 1077
+    },
+    {
+      "epoch": 0.3276844745041417,
+      "grad_norm": 3.337839126586914,
+      "learning_rate": 0.00015,
+      "loss": 4.580809116363525,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3279884489702865,
+      "grad_norm": 1.8693277835845947,
+      "learning_rate": 0.00015,
+      "loss": 4.186188697814941,
+      "step": 1079
+    },
+    {
+      "epoch": 0.32829242343643134,
+      "grad_norm": 1.937726378440857,
+      "learning_rate": 0.00015,
+      "loss": 4.452868461608887,
+      "step": 1080
+    },
+    {
+      "epoch": 0.32859639790257616,
+      "grad_norm": 2.0536398887634277,
+      "learning_rate": 0.00015,
+      "loss": 4.528855323791504,
+      "step": 1081
+    },
+    {
+      "epoch": 0.32890037236872105,
+      "grad_norm": 2.8802247047424316,
+      "learning_rate": 0.00015,
+      "loss": 4.407623291015625,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3292043468348659,
+      "grad_norm": 2.1634902954101562,
+      "learning_rate": 0.00015,
+      "loss": 4.219900131225586,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3295083213010107,
+      "grad_norm": 1.9650461673736572,
+      "learning_rate": 0.00015,
+      "loss": 4.481503486633301,
+      "step": 1084
+    },
+    {
+      "epoch": 0.32981229576715554,
+      "grad_norm": 2.7018003463745117,
+      "learning_rate": 0.00015,
+      "loss": 4.487911224365234,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3301162702333004,
+      "grad_norm": 1.763282060623169,
+      "learning_rate": 0.00015,
+      "loss": 4.349383354187012,
+      "step": 1086
+    },
+    {
+      "epoch": 0.33042024469944525,
+      "grad_norm": 2.281540632247925,
+      "learning_rate": 0.00015,
+      "loss": 4.46090030670166,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3307242191655901,
+      "grad_norm": 2.076044797897339,
+      "learning_rate": 0.00015,
+      "loss": 4.202225685119629,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3310281936317349,
+      "grad_norm": 2.5172905921936035,
+      "learning_rate": 0.00015,
+      "loss": 4.189465522766113,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3313321680978798,
+      "grad_norm": 2.3919599056243896,
+      "learning_rate": 0.00015,
+      "loss": 4.125179290771484,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3316361425640246,
+      "grad_norm": 2.505152940750122,
+      "learning_rate": 0.00015,
+      "loss": 4.605058670043945,
+      "step": 1091
+    },
+    {
+      "epoch": 0.33194011703016946,
+      "grad_norm": 1.8279204368591309,
+      "learning_rate": 0.00015,
+      "loss": 4.220155239105225,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3322440914963143,
+      "grad_norm": 1.8086042404174805,
+      "learning_rate": 0.00015,
+      "loss": 4.178547382354736,
+      "step": 1093
+    },
+    {
+      "epoch": 0.33254806596245917,
+      "grad_norm": 1.7796474695205688,
+      "learning_rate": 0.00015,
+      "loss": 4.3697052001953125,
+      "step": 1094
+    },
+    {
+      "epoch": 0.332852040428604,
+      "grad_norm": 1.8746988773345947,
+      "learning_rate": 0.00015,
+      "loss": 4.108738899230957,
+      "step": 1095
+    },
+    {
+      "epoch": 0.33315601489474883,
+      "grad_norm": 1.878389596939087,
+      "learning_rate": 0.00015,
+      "loss": 4.4542107582092285,
+      "step": 1096
+    },
+    {
+      "epoch": 0.33345998936089366,
+      "grad_norm": 1.81068754196167,
+      "learning_rate": 0.00015,
+      "loss": 4.586126327514648,
+      "step": 1097
+    },
+    {
+      "epoch": 0.33376396382703855,
+      "grad_norm": 1.8679008483886719,
+      "learning_rate": 0.00015,
+      "loss": 4.1875152587890625,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3340679382931834,
+      "grad_norm": 2.6923649311065674,
+      "learning_rate": 0.00015,
+      "loss": 4.39630126953125,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3343719127593282,
+      "grad_norm": 2.322003126144409,
+      "learning_rate": 0.00015,
+      "loss": 4.225685119628906,
+      "step": 1100
+    },
+    {
+      "epoch": 0.33467588722547303,
+      "grad_norm": 1.8528896570205688,
+      "learning_rate": 0.00015,
+      "loss": 4.182785511016846,
+      "step": 1101
+    },
+    {
+      "epoch": 0.3349798616916179,
+      "grad_norm": 1.9222912788391113,
+      "learning_rate": 0.00015,
+      "loss": 4.3679704666137695,
+      "step": 1102
+    },
+    {
+      "epoch": 0.33528383615776275,
+      "grad_norm": 1.762705683708191,
+      "learning_rate": 0.00015,
+      "loss": 4.27524471282959,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3355878106239076,
+      "grad_norm": 2.0277292728424072,
+      "learning_rate": 0.00015,
+      "loss": 4.562997817993164,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3358917850900524,
+      "grad_norm": 2.2395339012145996,
+      "learning_rate": 0.00015,
+      "loss": 4.528027534484863,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3361957595561973,
+      "grad_norm": 2.9747073650360107,
+      "learning_rate": 0.00015,
+      "loss": 4.429815769195557,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3364997340223421,
+      "grad_norm": 2.2827374935150146,
+      "learning_rate": 0.00015,
+      "loss": 4.177967071533203,
+      "step": 1107
+    },
+    {
+      "epoch": 0.33680370848848695,
+      "grad_norm": 2.230912208557129,
+      "learning_rate": 0.00015,
+      "loss": 4.47346830368042,
+      "step": 1108
+    },
+    {
+      "epoch": 0.33710768295463184,
+      "grad_norm": 21.155065536499023,
+      "learning_rate": 0.00015,
+      "loss": 4.226831912994385,
+      "step": 1109
+    },
+    {
+      "epoch": 0.33741165742077667,
+      "grad_norm": 2.1347272396087646,
+      "learning_rate": 0.00015,
+      "loss": 4.528800964355469,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3377156318869215,
+      "grad_norm": 2.257551670074463,
+      "learning_rate": 0.00015,
+      "loss": 4.0327301025390625,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3380196063530663,
+      "grad_norm": 1.8749741315841675,
+      "learning_rate": 0.00015,
+      "loss": 4.40855073928833,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3383235808192112,
+      "grad_norm": 1.8021701574325562,
+      "learning_rate": 0.00015,
+      "loss": 4.253387928009033,
+      "step": 1113
+    },
+    {
+      "epoch": 0.33862755528535604,
+      "grad_norm": 2.121856689453125,
+      "learning_rate": 0.00015,
+      "loss": 4.464548587799072,
+      "step": 1114
+    },
+    {
+      "epoch": 0.33893152975150087,
+      "grad_norm": 2.61667537689209,
+      "learning_rate": 0.00015,
+      "loss": 4.426912307739258,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3392355042176457,
+      "grad_norm": 2.2152318954467773,
+      "learning_rate": 0.00015,
+      "loss": 4.277388095855713,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3395394786837906,
+      "grad_norm": 1.972698450088501,
+      "learning_rate": 0.00015,
+      "loss": 4.17889928817749,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3398434531499354,
+      "grad_norm": 1.958216905593872,
+      "learning_rate": 0.00015,
+      "loss": 4.340052604675293,
+      "step": 1118
+    },
+    {
+      "epoch": 0.34014742761608024,
+      "grad_norm": 1.914201259613037,
+      "learning_rate": 0.00015,
+      "loss": 4.237743377685547,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3404514020822251,
+      "grad_norm": 2.0328705310821533,
+      "learning_rate": 0.00015,
+      "loss": 3.949950695037842,
+      "step": 1120
+    },
+    {
+      "epoch": 0.34075537654836996,
+      "grad_norm": 1.911206841468811,
+      "learning_rate": 0.00015,
+      "loss": 4.552268981933594,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3410593510145148,
+      "grad_norm": 2.5535056591033936,
+      "learning_rate": 0.00015,
+      "loss": 4.56083345413208,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3413633254806596,
+      "grad_norm": 2.028970956802368,
+      "learning_rate": 0.00015,
+      "loss": 4.181200981140137,
+      "step": 1123
+    },
+    {
+      "epoch": 0.34166729994680445,
+      "grad_norm": 1.9840599298477173,
+      "learning_rate": 0.00015,
+      "loss": 4.266005516052246,
+      "step": 1124
+    },
+    {
+      "epoch": 0.34197127441294933,
+      "grad_norm": 1.9153393507003784,
+      "learning_rate": 0.00015,
+      "loss": 4.281067848205566,
+      "step": 1125
+    },
+    {
+      "epoch": 0.34227524887909416,
+      "grad_norm": 1.7173157930374146,
+      "learning_rate": 0.00015,
+      "loss": 4.359532356262207,
+      "step": 1126
+    },
+    {
+      "epoch": 0.342579223345239,
+      "grad_norm": 1.8509682416915894,
+      "learning_rate": 0.00015,
+      "loss": 4.309632301330566,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3428831978113838,
+      "grad_norm": 3.028641939163208,
+      "learning_rate": 0.00015,
+      "loss": 4.358727931976318,
+      "step": 1128
+    },
+    {
+      "epoch": 0.3431871722775287,
+      "grad_norm": 2.0731918811798096,
+      "learning_rate": 0.00015,
+      "loss": 4.381097316741943,
+      "step": 1129
+    },
+    {
+      "epoch": 0.34349114674367354,
+      "grad_norm": 2.2629892826080322,
+      "learning_rate": 0.00015,
+      "loss": 4.089568138122559,
+      "step": 1130
+    },
+    {
+      "epoch": 0.34379512120981837,
+      "grad_norm": 1.9989763498306274,
+      "learning_rate": 0.00015,
+      "loss": 4.3608293533325195,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3440990956759632,
+      "grad_norm": 2.3224430084228516,
+      "learning_rate": 0.00015,
+      "loss": 4.255246639251709,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3444030701421081,
+      "grad_norm": 2.339310884475708,
+      "learning_rate": 0.00015,
+      "loss": 4.538227081298828,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3447070446082529,
+      "grad_norm": 1.9597009420394897,
+      "learning_rate": 0.00015,
+      "loss": 4.393087387084961,
+      "step": 1134
+    },
+    {
+      "epoch": 0.34501101907439774,
+      "grad_norm": 2.2850773334503174,
+      "learning_rate": 0.00015,
+      "loss": 4.572943687438965,
+      "step": 1135
+    },
+    {
+      "epoch": 0.34531499354054257,
+      "grad_norm": 4.200267314910889,
+      "learning_rate": 0.00015,
+      "loss": 4.071856498718262,
+      "step": 1136
+    },
+    {
+      "epoch": 0.34561896800668745,
+      "grad_norm": 2.442056179046631,
+      "learning_rate": 0.00015,
+      "loss": 4.1433610916137695,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3459229424728323,
+      "grad_norm": 1.9552514553070068,
+      "learning_rate": 0.00015,
+      "loss": 4.119884490966797,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3462269169389771,
+      "grad_norm": 2.0008952617645264,
+      "learning_rate": 0.00015,
+      "loss": 4.229976654052734,
+      "step": 1139
+    },
+    {
+      "epoch": 0.34653089140512194,
+      "grad_norm": 2.949995994567871,
+      "learning_rate": 0.00015,
+      "loss": 4.4145941734313965,
+      "step": 1140
+    },
+    {
+      "epoch": 0.34683486587126683,
+      "grad_norm": 2.2155494689941406,
+      "learning_rate": 0.00015,
+      "loss": 4.328131675720215,
+      "step": 1141
+    },
+    {
+      "epoch": 0.34713884033741166,
+      "grad_norm": 2.2986390590667725,
+      "learning_rate": 0.00015,
+      "loss": 4.207868576049805,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3474428148035565,
+      "grad_norm": 2.6639461517333984,
+      "learning_rate": 0.00015,
+      "loss": 4.404623985290527,
+      "step": 1143
+    },
+    {
+      "epoch": 0.3477467892697013,
+      "grad_norm": 1.8962833881378174,
+      "learning_rate": 0.00015,
+      "loss": 4.257451057434082,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3480507637358462,
+      "grad_norm": 1.8360495567321777,
+      "learning_rate": 0.00015,
+      "loss": 4.511965751647949,
+      "step": 1145
+    },
+    {
+      "epoch": 0.34835473820199103,
+      "grad_norm": 2.3360161781311035,
+      "learning_rate": 0.00015,
+      "loss": 4.3190016746521,
+      "step": 1146
+    },
+    {
+      "epoch": 0.34865871266813586,
+      "grad_norm": 2.683791160583496,
+      "learning_rate": 0.00015,
+      "loss": 4.027572154998779,
+      "step": 1147
+    },
+    {
+      "epoch": 0.34896268713428075,
+      "grad_norm": 1.8634052276611328,
+      "learning_rate": 0.00015,
+      "loss": 4.021484375,
+      "step": 1148
+    },
+    {
+      "epoch": 0.3492666616004256,
+      "grad_norm": 1.6122045516967773,
+      "learning_rate": 0.00015,
+      "loss": 4.197887897491455,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3495706360665704,
+      "grad_norm": 1.7859269380569458,
+      "learning_rate": 0.00015,
+      "loss": 4.3818511962890625,
+      "step": 1150
+    },
+    {
+      "epoch": 0.34987461053271524,
+      "grad_norm": 2.0781431198120117,
+      "learning_rate": 0.00015,
+      "loss": 4.144061088562012,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3501785849988601,
+      "grad_norm": 2.1974048614501953,
+      "learning_rate": 0.00015,
+      "loss": 4.208093166351318,
+      "step": 1152
+    },
+    {
+      "epoch": 0.35048255946500495,
+      "grad_norm": 1.9238510131835938,
+      "learning_rate": 0.00015,
+      "loss": 4.579778671264648,
+      "step": 1153
+    },
+    {
+      "epoch": 0.3507865339311498,
+      "grad_norm": 1.4841136932373047,
+      "learning_rate": 0.00015,
+      "loss": 4.534719467163086,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3510905083972946,
+      "grad_norm": 2.1653358936309814,
+      "learning_rate": 0.00015,
+      "loss": 4.164756774902344,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3513944828634395,
+      "grad_norm": 2.4431099891662598,
+      "learning_rate": 0.00015,
+      "loss": 4.932329177856445,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3516984573295843,
+      "grad_norm": 2.287158250808716,
+      "learning_rate": 0.00015,
+      "loss": 4.265947341918945,
+      "step": 1157
+    },
+    {
+      "epoch": 0.35200243179572915,
+      "grad_norm": 2.2220795154571533,
+      "learning_rate": 0.00015,
+      "loss": 4.132450103759766,
+      "step": 1158
+    },
+    {
+      "epoch": 0.352306406261874,
+      "grad_norm": 1.7991541624069214,
+      "learning_rate": 0.00015,
+      "loss": 4.622800827026367,
+      "step": 1159
+    },
+    {
+      "epoch": 0.35261038072801887,
+      "grad_norm": 2.20267391204834,
+      "learning_rate": 0.00015,
+      "loss": 4.4204607009887695,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3529143551941637,
+      "grad_norm": 1.823480486869812,
+      "learning_rate": 0.00015,
+      "loss": 4.123553276062012,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3532183296603085,
+      "grad_norm": 1.9210087060928345,
+      "learning_rate": 0.00015,
+      "loss": 4.105894565582275,
+      "step": 1162
+    },
+    {
+      "epoch": 0.35352230412645336,
+      "grad_norm": 2.881580114364624,
+      "learning_rate": 0.00015,
+      "loss": 4.572296142578125,
+      "step": 1163
+    },
+    {
+      "epoch": 0.35382627859259824,
+      "grad_norm": 1.804044485092163,
+      "learning_rate": 0.00015,
+      "loss": 4.289466857910156,
+      "step": 1164
+    },
+    {
+      "epoch": 0.35413025305874307,
+      "grad_norm": 2.4966917037963867,
+      "learning_rate": 0.00015,
+      "loss": 4.154740333557129,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3544342275248879,
+      "grad_norm": 2.0871667861938477,
+      "learning_rate": 0.00015,
+      "loss": 4.3722124099731445,
+      "step": 1166
+    },
+    {
+      "epoch": 0.35473820199103273,
+      "grad_norm": 2.059368371963501,
+      "learning_rate": 0.00015,
+      "loss": 4.468001365661621,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3550421764571776,
+      "grad_norm": 1.5784777402877808,
+      "learning_rate": 0.00015,
+      "loss": 4.475811958312988,
+      "step": 1168
+    },
+    {
+      "epoch": 0.35534615092332245,
+      "grad_norm": 2.280362844467163,
+      "learning_rate": 0.00015,
+      "loss": 4.334688186645508,
+      "step": 1169
+    },
+    {
+      "epoch": 0.3556501253894673,
+      "grad_norm": 2.8135180473327637,
+      "learning_rate": 0.00015,
+      "loss": 4.609158039093018,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3559540998556121,
+      "grad_norm": 2.0751149654388428,
+      "learning_rate": 0.00015,
+      "loss": 4.207970142364502,
+      "step": 1171
+    },
+    {
+      "epoch": 0.356258074321757,
+      "grad_norm": 2.148135185241699,
+      "learning_rate": 0.00015,
+      "loss": 4.384950637817383,
+      "step": 1172
+    },
+    {
+      "epoch": 0.3565620487879018,
+      "grad_norm": 1.940162181854248,
+      "learning_rate": 0.00015,
+      "loss": 4.276330471038818,
+      "step": 1173
+    },
+    {
+      "epoch": 0.35686602325404665,
+      "grad_norm": 1.6265183687210083,
+      "learning_rate": 0.00015,
+      "loss": 4.337113380432129,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3571699977201915,
+      "grad_norm": 2.144827127456665,
+      "learning_rate": 0.00015,
+      "loss": 4.30496883392334,
+      "step": 1175
+    },
+    {
+      "epoch": 0.35747397218633636,
+      "grad_norm": 1.8016746044158936,
+      "learning_rate": 0.00015,
+      "loss": 4.2445807456970215,
+      "step": 1176
+    },
+    {
+      "epoch": 0.3577779466524812,
+      "grad_norm": 1.9754371643066406,
+      "learning_rate": 0.00015,
+      "loss": 4.611670970916748,
+      "step": 1177
+    },
+    {
+      "epoch": 0.358081921118626,
+      "grad_norm": 5.0596699714660645,
+      "learning_rate": 0.00015,
+      "loss": 4.2333526611328125,
+      "step": 1178
+    },
+    {
+      "epoch": 0.35838589558477085,
+      "grad_norm": 2.066087245941162,
+      "learning_rate": 0.00015,
+      "loss": 4.183619499206543,
+      "step": 1179
+    },
+    {
+      "epoch": 0.35868987005091574,
+      "grad_norm": 1.997763991355896,
+      "learning_rate": 0.00015,
+      "loss": 4.462874889373779,
+      "step": 1180
+    },
+    {
+      "epoch": 0.35899384451706057,
+      "grad_norm": 2.1041746139526367,
+      "learning_rate": 0.00015,
+      "loss": 4.556066036224365,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3592978189832054,
+      "grad_norm": 2.523191452026367,
+      "learning_rate": 0.00015,
+      "loss": 4.241199493408203,
+      "step": 1182
+    },
+    {
+      "epoch": 0.3596017934493503,
+      "grad_norm": 2.196647882461548,
+      "learning_rate": 0.00015,
+      "loss": 4.389953136444092,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3599057679154951,
+      "grad_norm": 2.3794639110565186,
+      "learning_rate": 0.00015,
+      "loss": 4.341463088989258,
+      "step": 1184
+    },
+    {
+      "epoch": 0.36020974238163994,
+      "grad_norm": 2.2464449405670166,
+      "learning_rate": 0.00015,
+      "loss": 4.357426643371582,
+      "step": 1185
+    },
+    {
+      "epoch": 0.36051371684778477,
+      "grad_norm": 2.103778123855591,
+      "learning_rate": 0.00015,
+      "loss": 4.353619575500488,
+      "step": 1186
+    },
+    {
+      "epoch": 0.36081769131392966,
+      "grad_norm": 2.847456216812134,
+      "learning_rate": 0.00015,
+      "loss": 4.497030258178711,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3611216657800745,
+      "grad_norm": 2.4193923473358154,
+      "learning_rate": 0.00015,
+      "loss": 4.226492404937744,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3614256402462193,
+      "grad_norm": 1.7061328887939453,
+      "learning_rate": 0.00015,
+      "loss": 4.3179497718811035,
+      "step": 1189
+    },
+    {
+      "epoch": 0.36172961471236414,
+      "grad_norm": 1.5957392454147339,
+      "learning_rate": 0.00015,
+      "loss": 4.1068620681762695,
+      "step": 1190
+    },
+    {
+      "epoch": 0.36203358917850903,
+      "grad_norm": 2.1233084201812744,
+      "learning_rate": 0.00015,
+      "loss": 4.325814247131348,
+      "step": 1191
+    },
+    {
+      "epoch": 0.36233756364465386,
+      "grad_norm": 1.8084458112716675,
+      "learning_rate": 0.00015,
+      "loss": 4.472064018249512,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3626415381107987,
+      "grad_norm": 1.9206609725952148,
+      "learning_rate": 0.00015,
+      "loss": 4.137831687927246,
+      "step": 1193
+    },
+    {
+      "epoch": 0.3629455125769435,
+      "grad_norm": 1.8341199159622192,
+      "learning_rate": 0.00015,
+      "loss": 4.387076377868652,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3632494870430884,
+      "grad_norm": 1.7988574504852295,
+      "learning_rate": 0.00015,
+      "loss": 4.249260902404785,
+      "step": 1195
+    },
+    {
+      "epoch": 0.36355346150923323,
+      "grad_norm": 1.9418954849243164,
+      "learning_rate": 0.00015,
+      "loss": 4.326632499694824,
+      "step": 1196
+    },
+    {
+      "epoch": 0.36385743597537806,
+      "grad_norm": 2.1627252101898193,
+      "learning_rate": 0.00015,
+      "loss": 4.6474528312683105,
+      "step": 1197
+    },
+    {
+      "epoch": 0.3641614104415229,
+      "grad_norm": 1.9200024604797363,
+      "learning_rate": 0.00015,
+      "loss": 4.26380729675293,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3644653849076678,
+      "grad_norm": 1.9634867906570435,
+      "learning_rate": 0.00015,
+      "loss": 4.255786895751953,
+      "step": 1199
+    },
+    {
+      "epoch": 0.3647693593738126,
+      "grad_norm": 1.9351669549942017,
+      "learning_rate": 0.00015,
+      "loss": 4.1349616050720215,
+      "step": 1200
+    },
+    {
+      "epoch": 0.36507333383995744,
+      "grad_norm": 2.1029462814331055,
+      "learning_rate": 0.00015,
+      "loss": 4.121653079986572,
+      "step": 1201
+    },
+    {
+      "epoch": 0.36537730830610227,
+      "grad_norm": 1.9668861627578735,
+      "learning_rate": 0.00015,
+      "loss": 4.453611373901367,
+      "step": 1202
+    },
+    {
+      "epoch": 0.36568128277224715,
+      "grad_norm": 4.587677478790283,
+      "learning_rate": 0.00015,
+      "loss": 4.396587371826172,
+      "step": 1203
+    },
+    {
+      "epoch": 0.365985257238392,
+      "grad_norm": 1.9542491436004639,
+      "learning_rate": 0.00015,
+      "loss": 4.409153938293457,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3662892317045368,
+      "grad_norm": 2.102421760559082,
+      "learning_rate": 0.00015,
+      "loss": 4.175942897796631,
+      "step": 1205
+    },
+    {
+      "epoch": 0.36659320617068164,
+      "grad_norm": 1.7785810232162476,
+      "learning_rate": 0.00015,
+      "loss": 4.1811628341674805,
+      "step": 1206
+    },
+    {
+      "epoch": 0.3668971806368265,
+      "grad_norm": 1.6904630661010742,
+      "learning_rate": 0.00015,
+      "loss": 4.572501182556152,
+      "step": 1207
+    },
+    {
+      "epoch": 0.36720115510297135,
+      "grad_norm": 2.1081297397613525,
+      "learning_rate": 0.00015,
+      "loss": 4.4235944747924805,
+      "step": 1208
+    },
+    {
+      "epoch": 0.3675051295691162,
+      "grad_norm": 1.9664595127105713,
+      "learning_rate": 0.00015,
+      "loss": 4.154356956481934,
+      "step": 1209
+    },
+    {
+      "epoch": 0.367809104035261,
+      "grad_norm": 1.906436800956726,
+      "learning_rate": 0.00015,
+      "loss": 4.338773727416992,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3681130785014059,
+      "grad_norm": 1.858425259590149,
+      "learning_rate": 0.00015,
+      "loss": 4.478194713592529,
+      "step": 1211
+    },
+    {
+      "epoch": 0.36841705296755073,
+      "grad_norm": 2.040677309036255,
+      "learning_rate": 0.00015,
+      "loss": 4.254550933837891,
+      "step": 1212
+    },
+    {
+      "epoch": 0.36872102743369556,
+      "grad_norm": 2.2248830795288086,
+      "learning_rate": 0.00015,
+      "loss": 4.163328647613525,
+      "step": 1213
+    },
+    {
+      "epoch": 0.3690250018998404,
+      "grad_norm": 2.2339770793914795,
+      "learning_rate": 0.00015,
+      "loss": 4.425605297088623,
+      "step": 1214
+    },
+    {
+      "epoch": 0.3693289763659853,
+      "grad_norm": 2.05289626121521,
+      "learning_rate": 0.00015,
+      "loss": 4.325355529785156,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3696329508321301,
+      "grad_norm": 1.7779417037963867,
+      "learning_rate": 0.00015,
+      "loss": 4.218346118927002,
+      "step": 1216
+    },
+    {
+      "epoch": 0.36993692529827493,
+      "grad_norm": 1.8951629400253296,
+      "learning_rate": 0.00015,
+      "loss": 4.103420257568359,
+      "step": 1217
+    },
+    {
+      "epoch": 0.37024089976441976,
+      "grad_norm": 2.106891393661499,
+      "learning_rate": 0.00015,
+      "loss": 4.3794941902160645,
+      "step": 1218
+    },
+    {
+      "epoch": 0.37054487423056465,
+      "grad_norm": 1.8081761598587036,
+      "learning_rate": 0.00015,
+      "loss": 4.348600387573242,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3708488486967095,
+      "grad_norm": 1.8789769411087036,
+      "learning_rate": 0.00015,
+      "loss": 4.1857709884643555,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3711528231628543,
+      "grad_norm": 2.0007784366607666,
+      "learning_rate": 0.00015,
+      "loss": 4.058677673339844,
+      "step": 1221
+    },
+    {
+      "epoch": 0.3714567976289992,
+      "grad_norm": 3.0004665851593018,
+      "learning_rate": 0.00015,
+      "loss": 4.468400001525879,
+      "step": 1222
+    },
+    {
+      "epoch": 0.371760772095144,
+      "grad_norm": 1.995736002922058,
+      "learning_rate": 0.00015,
+      "loss": 4.295952796936035,
+      "step": 1223
+    },
+    {
+      "epoch": 0.37206474656128885,
+      "grad_norm": 2.054579019546509,
+      "learning_rate": 0.00015,
+      "loss": 4.629749298095703,
+      "step": 1224
+    },
+    {
+      "epoch": 0.3723687210274337,
+      "grad_norm": 2.1252148151397705,
+      "learning_rate": 0.00015,
+      "loss": 4.055646896362305,
+      "step": 1225
+    },
+    {
+      "epoch": 0.37267269549357857,
+      "grad_norm": 1.7454242706298828,
+      "learning_rate": 0.00015,
+      "loss": 4.187529563903809,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3729766699597234,
+      "grad_norm": 2.0207037925720215,
+      "learning_rate": 0.00015,
+      "loss": 4.666647911071777,
+      "step": 1227
+    },
+    {
+      "epoch": 0.3732806444258682,
+      "grad_norm": 2.2250475883483887,
+      "learning_rate": 0.00015,
+      "loss": 4.653911590576172,
+      "step": 1228
+    },
+    {
+      "epoch": 0.37358461889201305,
+      "grad_norm": 1.9888203144073486,
+      "learning_rate": 0.00015,
+      "loss": 4.246804714202881,
+      "step": 1229
+    },
+    {
+      "epoch": 0.37388859335815794,
+      "grad_norm": 2.6151084899902344,
+      "learning_rate": 0.00015,
+      "loss": 4.556540489196777,
+      "step": 1230
+    },
+    {
+      "epoch": 0.37419256782430277,
+      "grad_norm": 1.8919669389724731,
+      "learning_rate": 0.00015,
+      "loss": 4.19811487197876,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3744965422904476,
+      "grad_norm": 1.7876086235046387,
+      "learning_rate": 0.00015,
+      "loss": 4.141073703765869,
+      "step": 1232
+    },
+    {
+      "epoch": 0.37480051675659243,
+      "grad_norm": 1.8092737197875977,
+      "learning_rate": 0.00015,
+      "loss": 4.333150863647461,
+      "step": 1233
+    },
+    {
+      "epoch": 0.3751044912227373,
+      "grad_norm": 2.859631061553955,
+      "learning_rate": 0.00015,
+      "loss": 4.249753952026367,
+      "step": 1234
+    },
+    {
+      "epoch": 0.37540846568888214,
+      "grad_norm": 1.7699271440505981,
+      "learning_rate": 0.00015,
+      "loss": 4.0793561935424805,
+      "step": 1235
+    },
+    {
+      "epoch": 0.37571244015502697,
+      "grad_norm": 1.888400673866272,
+      "learning_rate": 0.00015,
+      "loss": 4.420899391174316,
+      "step": 1236
+    },
+    {
+      "epoch": 0.3760164146211718,
+      "grad_norm": 1.7865756750106812,
+      "learning_rate": 0.00015,
+      "loss": 3.9361071586608887,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3763203890873167,
+      "grad_norm": 2.100719451904297,
+      "learning_rate": 0.00015,
+      "loss": 4.285518646240234,
+      "step": 1238
+    },
+    {
+      "epoch": 0.3766243635534615,
+      "grad_norm": 1.9925858974456787,
+      "learning_rate": 0.00015,
+      "loss": 4.253064155578613,
+      "step": 1239
+    },
+    {
+      "epoch": 0.37692833801960635,
+      "grad_norm": 1.8425103425979614,
+      "learning_rate": 0.00015,
+      "loss": 4.2561798095703125,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3772323124857512,
+      "grad_norm": 2.0761773586273193,
+      "learning_rate": 0.00015,
+      "loss": 4.329405307769775,
+      "step": 1241
+    },
+    {
+      "epoch": 0.37753628695189606,
+      "grad_norm": 3.2859809398651123,
+      "learning_rate": 0.00015,
+      "loss": 4.515865802764893,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3778402614180409,
+      "grad_norm": 1.9178838729858398,
+      "learning_rate": 0.00015,
+      "loss": 4.294964790344238,
+      "step": 1243
+    },
+    {
+      "epoch": 0.3781442358841857,
+      "grad_norm": 1.7650163173675537,
+      "learning_rate": 0.00015,
+      "loss": 4.389337062835693,
+      "step": 1244
+    },
+    {
+      "epoch": 0.37844821035033055,
+      "grad_norm": 1.9344691038131714,
+      "learning_rate": 0.00015,
+      "loss": 4.056518077850342,
+      "step": 1245
+    },
+    {
+      "epoch": 0.37875218481647543,
+      "grad_norm": 3.2899160385131836,
+      "learning_rate": 0.00015,
+      "loss": 4.434074401855469,
+      "step": 1246
+    },
+    {
+      "epoch": 0.37905615928262026,
+      "grad_norm": 1.9324767589569092,
+      "learning_rate": 0.00015,
+      "loss": 4.220419883728027,
+      "step": 1247
+    },
+    {
+      "epoch": 0.3793601337487651,
+      "grad_norm": 1.6367474794387817,
+      "learning_rate": 0.00015,
+      "loss": 4.101804733276367,
+      "step": 1248
+    },
+    {
+      "epoch": 0.3796641082149099,
+      "grad_norm": 1.8838940858840942,
+      "learning_rate": 0.00015,
+      "loss": 4.274724006652832,
+      "step": 1249
+    },
+    {
+      "epoch": 0.3799680826810548,
+      "grad_norm": 2.459381103515625,
+      "learning_rate": 0.00015,
+      "loss": 4.18360710144043,
+      "step": 1250
+    },
+    {
+      "epoch": 0.38027205714719964,
+      "grad_norm": 1.9690313339233398,
+      "learning_rate": 0.00015,
+      "loss": 4.190372467041016,
+      "step": 1251
+    },
+    {
+      "epoch": 0.38057603161334447,
+      "grad_norm": 1.9526063203811646,
+      "learning_rate": 0.00015,
+      "loss": 4.288544178009033,
+      "step": 1252
+    },
+    {
+      "epoch": 0.3808800060794893,
+      "grad_norm": 1.953754186630249,
+      "learning_rate": 0.00015,
+      "loss": 4.216185569763184,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3811839805456342,
+      "grad_norm": 1.9070219993591309,
+      "learning_rate": 0.00015,
+      "loss": 4.285344123840332,
+      "step": 1254
+    },
+    {
+      "epoch": 0.381487955011779,
+      "grad_norm": 3.7306220531463623,
+      "learning_rate": 0.00015,
+      "loss": 4.083676815032959,
+      "step": 1255
+    },
+    {
+      "epoch": 0.38179192947792384,
+      "grad_norm": 2.248500108718872,
+      "learning_rate": 0.00015,
+      "loss": 4.225886344909668,
+      "step": 1256
+    },
+    {
+      "epoch": 0.38209590394406867,
+      "grad_norm": 1.7171469926834106,
+      "learning_rate": 0.00015,
+      "loss": 4.347973823547363,
+      "step": 1257
+    },
+    {
+      "epoch": 0.38239987841021356,
+      "grad_norm": 1.8810170888900757,
+      "learning_rate": 0.00015,
+      "loss": 4.151370525360107,
+      "step": 1258
+    },
+    {
+      "epoch": 0.3827038528763584,
+      "grad_norm": 1.8285266160964966,
+      "learning_rate": 0.00015,
+      "loss": 4.26771354675293,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3830078273425032,
+      "grad_norm": 1.9770228862762451,
+      "learning_rate": 0.00015,
+      "loss": 4.268593788146973,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3833118018086481,
+      "grad_norm": 2.1152875423431396,
+      "learning_rate": 0.00015,
+      "loss": 4.164984226226807,
+      "step": 1261
+    },
+    {
+      "epoch": 0.38361577627479293,
+      "grad_norm": 1.8868613243103027,
+      "learning_rate": 0.00015,
+      "loss": 4.336614608764648,
+      "step": 1262
+    },
+    {
+      "epoch": 0.38391975074093776,
+      "grad_norm": 1.9492988586425781,
+      "learning_rate": 0.00015,
+      "loss": 4.2262797355651855,
+      "step": 1263
+    },
+    {
+      "epoch": 0.3842237252070826,
+      "grad_norm": 2.4051339626312256,
+      "learning_rate": 0.00015,
+      "loss": 4.236055374145508,
+      "step": 1264
+    },
+    {
+      "epoch": 0.3845276996732275,
+      "grad_norm": 1.8402055501937866,
+      "learning_rate": 0.00015,
+      "loss": 4.151636123657227,
+      "step": 1265
+    },
+    {
+      "epoch": 0.3848316741393723,
+      "grad_norm": 1.937568187713623,
+      "learning_rate": 0.00015,
+      "loss": 4.377797603607178,
+      "step": 1266
+    },
+    {
+      "epoch": 0.38513564860551713,
+      "grad_norm": 1.8739315271377563,
+      "learning_rate": 0.00015,
+      "loss": 4.021393775939941,
+      "step": 1267
+    },
+    {
+      "epoch": 0.38543962307166196,
+      "grad_norm": 1.7112767696380615,
+      "learning_rate": 0.00015,
+      "loss": 4.262368202209473,
+      "step": 1268
+    },
+    {
+      "epoch": 0.38574359753780685,
+      "grad_norm": 2.432769298553467,
+      "learning_rate": 0.00015,
+      "loss": 3.75838565826416,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3860475720039517,
+      "grad_norm": 2.2427845001220703,
+      "learning_rate": 0.00015,
+      "loss": 4.119418621063232,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3863515464700965,
+      "grad_norm": 2.0331101417541504,
+      "learning_rate": 0.00015,
+      "loss": 4.012245178222656,
+      "step": 1271
+    },
+    {
+      "epoch": 0.38665552093624134,
+      "grad_norm": 2.512636661529541,
+      "learning_rate": 0.00015,
+      "loss": 3.8470547199249268,
+      "step": 1272
+    },
+    {
+      "epoch": 0.3869594954023862,
+      "grad_norm": 1.7639203071594238,
+      "learning_rate": 0.00015,
+      "loss": 4.253369331359863,
+      "step": 1273
+    },
+    {
+      "epoch": 0.38726346986853105,
+      "grad_norm": 2.221543073654175,
+      "learning_rate": 0.00015,
+      "loss": 4.323329925537109,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3875674443346759,
+      "grad_norm": 2.5534448623657227,
+      "learning_rate": 0.00015,
+      "loss": 4.398746490478516,
+      "step": 1275
+    },
+    {
+      "epoch": 0.3878714188008207,
+      "grad_norm": 2.4140677452087402,
+      "learning_rate": 0.00015,
+      "loss": 4.505166053771973,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3881753932669656,
+      "grad_norm": 2.655910015106201,
+      "learning_rate": 0.00015,
+      "loss": 4.128080368041992,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3884793677331104,
+      "grad_norm": 1.7601017951965332,
+      "learning_rate": 0.00015,
+      "loss": 4.109560966491699,
+      "step": 1278
+    },
+    {
+      "epoch": 0.38878334219925526,
+      "grad_norm": 1.9833213090896606,
+      "learning_rate": 0.00015,
+      "loss": 4.195367813110352,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3890873166654001,
+      "grad_norm": 2.197103261947632,
+      "learning_rate": 0.00015,
+      "loss": 4.330955505371094,
+      "step": 1280
+    },
+    {
+      "epoch": 0.38939129113154497,
+      "grad_norm": 2.0471062660217285,
+      "learning_rate": 0.00015,
+      "loss": 4.059005260467529,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3896952655976898,
+      "grad_norm": 2.2260994911193848,
+      "learning_rate": 0.00015,
+      "loss": 4.279693603515625,
+      "step": 1282
+    },
+    {
+      "epoch": 0.38999924006383463,
+      "grad_norm": 2.042323589324951,
+      "learning_rate": 0.00015,
+      "loss": 4.136223316192627,
+      "step": 1283
+    },
+    {
+      "epoch": 0.39030321452997946,
+      "grad_norm": 1.7412962913513184,
+      "learning_rate": 0.00015,
+      "loss": 4.201234817504883,
+      "step": 1284
+    },
+    {
+      "epoch": 0.39060718899612434,
+      "grad_norm": 1.9684971570968628,
+      "learning_rate": 0.00015,
+      "loss": 4.118769645690918,
+      "step": 1285
+    },
+    {
+      "epoch": 0.3909111634622692,
+      "grad_norm": 2.1194043159484863,
+      "learning_rate": 0.00015,
+      "loss": 4.384788513183594,
+      "step": 1286
+    },
+    {
+      "epoch": 0.391215137928414,
+      "grad_norm": 1.9592872858047485,
+      "learning_rate": 0.00015,
+      "loss": 4.117946624755859,
+      "step": 1287
+    },
+    {
+      "epoch": 0.39151911239455883,
+      "grad_norm": 2.4148428440093994,
+      "learning_rate": 0.00015,
+      "loss": 4.202004432678223,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3918230868607037,
+      "grad_norm": 2.7596030235290527,
+      "learning_rate": 0.00015,
+      "loss": 4.105706691741943,
+      "step": 1289
+    },
+    {
+      "epoch": 0.39212706132684855,
+      "grad_norm": 2.0022737979888916,
+      "learning_rate": 0.00015,
+      "loss": 4.460059642791748,
+      "step": 1290
+    },
+    {
+      "epoch": 0.3924310357929934,
+      "grad_norm": 2.4023711681365967,
+      "learning_rate": 0.00015,
+      "loss": 4.4527058601379395,
+      "step": 1291
+    },
+    {
+      "epoch": 0.3927350102591382,
+      "grad_norm": 1.9343289136886597,
+      "learning_rate": 0.00015,
+      "loss": 4.51945686340332,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3930389847252831,
+      "grad_norm": 2.1259825229644775,
+      "learning_rate": 0.00015,
+      "loss": 4.148380279541016,
+      "step": 1293
+    },
+    {
+      "epoch": 0.3933429591914279,
+      "grad_norm": 2.1968870162963867,
+      "learning_rate": 0.00015,
+      "loss": 4.473582744598389,
+      "step": 1294
+    },
+    {
+      "epoch": 0.39364693365757275,
+      "grad_norm": 1.873962640762329,
+      "learning_rate": 0.00015,
+      "loss": 4.355568885803223,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3939509081237176,
+      "grad_norm": 1.6938269138336182,
+      "learning_rate": 0.00015,
+      "loss": 4.261201858520508,
+      "step": 1296
+    },
+    {
+      "epoch": 0.39425488258986247,
+      "grad_norm": 1.990830659866333,
+      "learning_rate": 0.00015,
+      "loss": 4.139786243438721,
+      "step": 1297
+    },
+    {
+      "epoch": 0.3945588570560073,
+      "grad_norm": 1.8179450035095215,
+      "learning_rate": 0.00015,
+      "loss": 4.038569927215576,
+      "step": 1298
+    },
+    {
+      "epoch": 0.3948628315221521,
+      "grad_norm": 1.8301723003387451,
+      "learning_rate": 0.00015,
+      "loss": 4.050404071807861,
+      "step": 1299
+    },
+    {
+      "epoch": 0.395166805988297,
+      "grad_norm": 1.922707200050354,
+      "learning_rate": 0.00015,
+      "loss": 4.506591320037842,
+      "step": 1300
+    },
+    {
+      "epoch": 0.39547078045444184,
+      "grad_norm": 2.305636405944824,
+      "learning_rate": 0.00015,
+      "loss": 3.988231658935547,
+      "step": 1301
+    },
+    {
+      "epoch": 0.39577475492058667,
+      "grad_norm": 2.124340295791626,
+      "learning_rate": 0.00015,
+      "loss": 4.318326950073242,
+      "step": 1302
+    },
+    {
+      "epoch": 0.3960787293867315,
+      "grad_norm": 2.1424598693847656,
+      "learning_rate": 0.00015,
+      "loss": 4.009394645690918,
+      "step": 1303
+    },
+    {
+      "epoch": 0.3963827038528764,
+      "grad_norm": 2.2413320541381836,
+      "learning_rate": 0.00015,
+      "loss": 4.2968854904174805,
+      "step": 1304
+    },
+    {
+      "epoch": 0.3966866783190212,
+      "grad_norm": 1.6211597919464111,
+      "learning_rate": 0.00015,
+      "loss": 4.26740837097168,
+      "step": 1305
+    },
+    {
+      "epoch": 0.39699065278516604,
+      "grad_norm": 1.9775257110595703,
+      "learning_rate": 0.00015,
+      "loss": 4.094832420349121,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3972946272513109,
+      "grad_norm": 2.3853938579559326,
+      "learning_rate": 0.00015,
+      "loss": 4.095998764038086,
+      "step": 1307
+    },
+    {
+      "epoch": 0.39759860171745576,
+      "grad_norm": 3.0847108364105225,
+      "learning_rate": 0.00015,
+      "loss": 4.547788619995117,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3979025761836006,
+      "grad_norm": 2.523942470550537,
+      "learning_rate": 0.00015,
+      "loss": 4.4586567878723145,
+      "step": 1309
+    },
+    {
+      "epoch": 0.3982065506497454,
+      "grad_norm": 2.005984306335449,
+      "learning_rate": 0.00015,
+      "loss": 4.074298858642578,
+      "step": 1310
+    },
+    {
+      "epoch": 0.39851052511589025,
+      "grad_norm": 1.9850512742996216,
+      "learning_rate": 0.00015,
+      "loss": 4.208410263061523,
+      "step": 1311
+    },
+    {
+      "epoch": 0.39881449958203513,
+      "grad_norm": 1.8494230508804321,
+      "learning_rate": 0.00015,
+      "loss": 4.454097270965576,
+      "step": 1312
+    },
+    {
+      "epoch": 0.39911847404817996,
+      "grad_norm": 3.3678054809570312,
+      "learning_rate": 0.00015,
+      "loss": 4.212475776672363,
+      "step": 1313
+    },
+    {
+      "epoch": 0.3994224485143248,
+      "grad_norm": 2.0187344551086426,
+      "learning_rate": 0.00015,
+      "loss": 4.0936970710754395,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3997264229804696,
+      "grad_norm": 2.18410587310791,
+      "learning_rate": 0.00015,
+      "loss": 4.1955366134643555,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4000303974466145,
+      "grad_norm": 2.869882345199585,
+      "learning_rate": 0.00015,
+      "loss": 4.324729919433594,
+      "step": 1316
+    },
+    {
+      "epoch": 0.40033437191275933,
+      "grad_norm": 2.3366928100585938,
+      "learning_rate": 0.00015,
+      "loss": 4.334259986877441,
+      "step": 1317
+    },
+    {
+      "epoch": 0.40063834637890416,
+      "grad_norm": 1.9355032444000244,
+      "learning_rate": 0.00015,
+      "loss": 4.2698163986206055,
+      "step": 1318
+    },
+    {
+      "epoch": 0.400942320845049,
+      "grad_norm": 2.4210877418518066,
+      "learning_rate": 0.00015,
+      "loss": 4.11591100692749,
+      "step": 1319
+    },
+    {
+      "epoch": 0.4012462953111939,
+      "grad_norm": 1.9671308994293213,
+      "learning_rate": 0.00015,
+      "loss": 4.277069091796875,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4015502697773387,
+      "grad_norm": 2.874530553817749,
+      "learning_rate": 0.00015,
+      "loss": 4.436254978179932,
+      "step": 1321
+    },
+    {
+      "epoch": 0.40185424424348354,
+      "grad_norm": 2.1109981536865234,
+      "learning_rate": 0.00015,
+      "loss": 3.950303792953491,
+      "step": 1322
+    },
+    {
+      "epoch": 0.40215821870962837,
+      "grad_norm": 1.9734985828399658,
+      "learning_rate": 0.00015,
+      "loss": 4.3122687339782715,
+      "step": 1323
+    },
+    {
+      "epoch": 0.40246219317577325,
+      "grad_norm": 2.171161651611328,
+      "learning_rate": 0.00015,
+      "loss": 4.418011665344238,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4027661676419181,
+      "grad_norm": 1.793000340461731,
+      "learning_rate": 0.00015,
+      "loss": 4.115985870361328,
+      "step": 1325
+    },
+    {
+      "epoch": 0.4030701421080629,
+      "grad_norm": 2.6729133129119873,
+      "learning_rate": 0.00015,
+      "loss": 4.062838554382324,
+      "step": 1326
+    },
+    {
+      "epoch": 0.40337411657420774,
+      "grad_norm": 3.012962579727173,
+      "learning_rate": 0.00015,
+      "loss": 4.258819103240967,
+      "step": 1327
+    },
+    {
+      "epoch": 0.4036780910403526,
+      "grad_norm": 1.8988255262374878,
+      "learning_rate": 0.00015,
+      "loss": 4.235583305358887,
+      "step": 1328
+    },
+    {
+      "epoch": 0.40398206550649746,
+      "grad_norm": 2.047008991241455,
+      "learning_rate": 0.00015,
+      "loss": 4.029030799865723,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4042860399726423,
+      "grad_norm": 2.142446279525757,
+      "learning_rate": 0.00015,
+      "loss": 3.9635190963745117,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4045900144387871,
+      "grad_norm": 2.100599765777588,
+      "learning_rate": 0.00015,
+      "loss": 4.257514953613281,
+      "step": 1331
+    },
+    {
+      "epoch": 0.404893988904932,
+      "grad_norm": 1.866510033607483,
+      "learning_rate": 0.00015,
+      "loss": 4.128789901733398,
+      "step": 1332
+    },
+    {
+      "epoch": 0.40519796337107683,
+      "grad_norm": 1.744545817375183,
+      "learning_rate": 0.00015,
+      "loss": 4.148487091064453,
+      "step": 1333
+    },
+    {
+      "epoch": 0.40550193783722166,
+      "grad_norm": 2.56496524810791,
+      "learning_rate": 0.00015,
+      "loss": 4.316514015197754,
+      "step": 1334
+    },
+    {
+      "epoch": 0.40580591230336654,
+      "grad_norm": 1.9824632406234741,
+      "learning_rate": 0.00015,
+      "loss": 4.151178359985352,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4061098867695114,
+      "grad_norm": 1.87270987033844,
+      "learning_rate": 0.00015,
+      "loss": 4.040594577789307,
+      "step": 1336
+    },
+    {
+      "epoch": 0.4064138612356562,
+      "grad_norm": 1.8191198110580444,
+      "learning_rate": 0.00015,
+      "loss": 3.9348669052124023,
+      "step": 1337
+    },
+    {
+      "epoch": 0.40671783570180103,
+      "grad_norm": 1.8206355571746826,
+      "learning_rate": 0.00015,
+      "loss": 4.670298099517822,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4070218101679459,
+      "grad_norm": 2.1933693885803223,
+      "learning_rate": 0.00015,
+      "loss": 3.972836971282959,
+      "step": 1339
+    },
+    {
+      "epoch": 0.40732578463409075,
+      "grad_norm": 2.0804970264434814,
+      "learning_rate": 0.00015,
+      "loss": 4.334033966064453,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4076297591002356,
+      "grad_norm": 2.5250837802886963,
+      "learning_rate": 0.00015,
+      "loss": 4.196231365203857,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4079337335663804,
+      "grad_norm": 2.172342538833618,
+      "learning_rate": 0.00015,
+      "loss": 3.974255084991455,
+      "step": 1342
+    },
+    {
+      "epoch": 0.4082377080325253,
+      "grad_norm": 3.174631118774414,
+      "learning_rate": 0.00015,
+      "loss": 3.9852652549743652,
+      "step": 1343
+    },
+    {
+      "epoch": 0.4085416824986701,
+      "grad_norm": 2.1556739807128906,
+      "learning_rate": 0.00015,
+      "loss": 4.26572847366333,
+      "step": 1344
+    },
+    {
+      "epoch": 0.40884565696481495,
+      "grad_norm": 2.278533458709717,
+      "learning_rate": 0.00015,
+      "loss": 4.36238956451416,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4091496314309598,
+      "grad_norm": 1.8759124279022217,
+      "learning_rate": 0.00015,
+      "loss": 4.200592994689941,
+      "step": 1346
+    },
+    {
+      "epoch": 0.40945360589710467,
+      "grad_norm": 1.9427827596664429,
+      "learning_rate": 0.00015,
+      "loss": 4.129970073699951,
+      "step": 1347
+    },
+    {
+      "epoch": 0.4097575803632495,
+      "grad_norm": 2.095383882522583,
+      "learning_rate": 0.00015,
+      "loss": 4.3097357749938965,
+      "step": 1348
+    },
+    {
+      "epoch": 0.4100615548293943,
+      "grad_norm": 2.1512913703918457,
+      "learning_rate": 0.00015,
+      "loss": 4.287084579467773,
+      "step": 1349
+    },
+    {
+      "epoch": 0.41036552929553916,
+      "grad_norm": 2.4184858798980713,
+      "learning_rate": 0.00015,
+      "loss": 4.458749771118164,
+      "step": 1350
+    },
+    {
+      "epoch": 0.41066950376168404,
+      "grad_norm": 2.4754788875579834,
+      "learning_rate": 0.00015,
+      "loss": 4.384786605834961,
+      "step": 1351
+    },
+    {
+      "epoch": 0.41097347822782887,
+      "grad_norm": 2.116229295730591,
+      "learning_rate": 0.00015,
+      "loss": 4.537047863006592,
+      "step": 1352
+    },
+    {
+      "epoch": 0.4112774526939737,
+      "grad_norm": 1.835519790649414,
+      "learning_rate": 0.00015,
+      "loss": 4.257972240447998,
+      "step": 1353
+    },
+    {
+      "epoch": 0.41158142716011853,
+      "grad_norm": 1.9529637098312378,
+      "learning_rate": 0.00015,
+      "loss": 4.085052013397217,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4118854016262634,
+      "grad_norm": 2.115342378616333,
+      "learning_rate": 0.00015,
+      "loss": 4.201634407043457,
+      "step": 1355
+    },
+    {
+      "epoch": 0.41218937609240824,
+      "grad_norm": 1.946313738822937,
+      "learning_rate": 0.00015,
+      "loss": 4.305756092071533,
+      "step": 1356
+    },
+    {
+      "epoch": 0.4124933505585531,
+      "grad_norm": 1.8691147565841675,
+      "learning_rate": 0.00015,
+      "loss": 4.1919026374816895,
+      "step": 1357
+    },
+    {
+      "epoch": 0.4127973250246979,
+      "grad_norm": 2.4796009063720703,
+      "learning_rate": 0.00015,
+      "loss": 4.162874221801758,
+      "step": 1358
+    },
+    {
+      "epoch": 0.4131012994908428,
+      "grad_norm": 2.420008659362793,
+      "learning_rate": 0.00015,
+      "loss": 4.154208183288574,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4134052739569876,
+      "grad_norm": 1.8740540742874146,
+      "learning_rate": 0.00015,
+      "loss": 4.536676406860352,
+      "step": 1360
+    },
+    {
+      "epoch": 0.41370924842313245,
+      "grad_norm": 1.8773142099380493,
+      "learning_rate": 0.00015,
+      "loss": 4.255711555480957,
+      "step": 1361
+    },
+    {
+      "epoch": 0.4140132228892773,
+      "grad_norm": 2.159271001815796,
+      "learning_rate": 0.00015,
+      "loss": 4.6029253005981445,
+      "step": 1362
+    },
+    {
+      "epoch": 0.41431719735542216,
+      "grad_norm": 1.8335107564926147,
+      "learning_rate": 0.00015,
+      "loss": 4.297633647918701,
+      "step": 1363
+    },
+    {
+      "epoch": 0.414621171821567,
+      "grad_norm": 1.9928083419799805,
+      "learning_rate": 0.00015,
+      "loss": 4.088753700256348,
+      "step": 1364
+    },
+    {
+      "epoch": 0.4149251462877118,
+      "grad_norm": 2.360952615737915,
+      "learning_rate": 0.00015,
+      "loss": 4.903768539428711,
+      "step": 1365
+    },
+    {
+      "epoch": 0.41522912075385665,
+      "grad_norm": 1.5310916900634766,
+      "learning_rate": 0.00015,
+      "loss": 4.273285865783691,
+      "step": 1366
+    },
+    {
+      "epoch": 0.41553309522000154,
+      "grad_norm": 1.8013471364974976,
+      "learning_rate": 0.00015,
+      "loss": 4.125875473022461,
+      "step": 1367
+    },
+    {
+      "epoch": 0.41583706968614637,
+      "grad_norm": 1.727344274520874,
+      "learning_rate": 0.00015,
+      "loss": 4.069706439971924,
+      "step": 1368
+    },
+    {
+      "epoch": 0.4161410441522912,
+      "grad_norm": 2.335592031478882,
+      "learning_rate": 0.00015,
+      "loss": 4.1445631980896,
+      "step": 1369
+    },
+    {
+      "epoch": 0.416445018618436,
+      "grad_norm": 2.093949794769287,
+      "learning_rate": 0.00015,
+      "loss": 4.518946647644043,
+      "step": 1370
+    },
+    {
+      "epoch": 0.4167489930845809,
+      "grad_norm": 1.9838320016860962,
+      "learning_rate": 0.00015,
+      "loss": 4.007177829742432,
+      "step": 1371
+    },
+    {
+      "epoch": 0.41705296755072574,
+      "grad_norm": 2.515556812286377,
+      "learning_rate": 0.00015,
+      "loss": 4.424863815307617,
+      "step": 1372
+    },
+    {
+      "epoch": 0.41735694201687057,
+      "grad_norm": 1.851576805114746,
+      "learning_rate": 0.00015,
+      "loss": 3.809074640274048,
+      "step": 1373
+    },
+    {
+      "epoch": 0.41766091648301545,
+      "grad_norm": 1.9523646831512451,
+      "learning_rate": 0.00015,
+      "loss": 3.986060619354248,
+      "step": 1374
+    },
+    {
+      "epoch": 0.4179648909491603,
+      "grad_norm": 2.3519034385681152,
+      "learning_rate": 0.00015,
+      "loss": 4.279633045196533,
+      "step": 1375
+    },
+    {
+      "epoch": 0.4182688654153051,
+      "grad_norm": 3.0041215419769287,
+      "learning_rate": 0.00015,
+      "loss": 4.247065544128418,
+      "step": 1376
+    },
+    {
+      "epoch": 0.41857283988144994,
+      "grad_norm": 2.2251083850860596,
+      "learning_rate": 0.00015,
+      "loss": 4.540241718292236,
+      "step": 1377
+    },
+    {
+      "epoch": 0.41887681434759483,
+      "grad_norm": 2.0129141807556152,
+      "learning_rate": 0.00015,
+      "loss": 4.433598518371582,
+      "step": 1378
+    },
+    {
+      "epoch": 0.41918078881373966,
+      "grad_norm": 1.9230022430419922,
+      "learning_rate": 0.00015,
+      "loss": 3.85391902923584,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4194847632798845,
+      "grad_norm": 1.948446273803711,
+      "learning_rate": 0.00015,
+      "loss": 4.42063570022583,
+      "step": 1380
+    },
+    {
+      "epoch": 0.4197887377460293,
+      "grad_norm": 2.624051809310913,
+      "learning_rate": 0.00015,
+      "loss": 4.053753852844238,
+      "step": 1381
+    },
+    {
+      "epoch": 0.4200927122121742,
+      "grad_norm": 2.5234336853027344,
+      "learning_rate": 0.00015,
+      "loss": 4.08837890625,
+      "step": 1382
+    },
+    {
+      "epoch": 0.42039668667831903,
+      "grad_norm": 2.132741689682007,
+      "learning_rate": 0.00015,
+      "loss": 3.969421863555908,
+      "step": 1383
+    },
+    {
+      "epoch": 0.42070066114446386,
+      "grad_norm": 1.7476507425308228,
+      "learning_rate": 0.00015,
+      "loss": 4.235538005828857,
+      "step": 1384
+    },
+    {
+      "epoch": 0.4210046356106087,
+      "grad_norm": 2.0522043704986572,
+      "learning_rate": 0.00015,
+      "loss": 4.6777825355529785,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4213086100767536,
+      "grad_norm": 1.850834608078003,
+      "learning_rate": 0.00015,
+      "loss": 4.130456924438477,
+      "step": 1386
+    },
+    {
+      "epoch": 0.4216125845428984,
+      "grad_norm": 1.7363988161087036,
+      "learning_rate": 0.00015,
+      "loss": 4.325787544250488,
+      "step": 1387
+    },
+    {
+      "epoch": 0.42191655900904323,
+      "grad_norm": 5.023498058319092,
+      "learning_rate": 0.00015,
+      "loss": 4.588955402374268,
+      "step": 1388
+    },
+    {
+      "epoch": 0.42222053347518806,
+      "grad_norm": 2.215956449508667,
+      "learning_rate": 0.00015,
+      "loss": 4.251990795135498,
+      "step": 1389
+    },
+    {
+      "epoch": 0.42252450794133295,
+      "grad_norm": 2.502789258956909,
+      "learning_rate": 0.00015,
+      "loss": 4.318276882171631,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4228284824074778,
+      "grad_norm": 2.2021732330322266,
+      "learning_rate": 0.00015,
+      "loss": 4.389657497406006,
+      "step": 1391
+    },
+    {
+      "epoch": 0.4231324568736226,
+      "grad_norm": 2.80069899559021,
+      "learning_rate": 0.00015,
+      "loss": 4.283598899841309,
+      "step": 1392
+    },
+    {
+      "epoch": 0.42343643133976744,
+      "grad_norm": 2.0096654891967773,
+      "learning_rate": 0.00015,
+      "loss": 4.291179656982422,
+      "step": 1393
+    },
+    {
+      "epoch": 0.4237404058059123,
+      "grad_norm": 2.932191848754883,
+      "learning_rate": 0.00015,
+      "loss": 4.280506610870361,
+      "step": 1394
+    },
+    {
+      "epoch": 0.42404438027205715,
+      "grad_norm": 2.0722994804382324,
+      "learning_rate": 0.00015,
+      "loss": 4.131236553192139,
+      "step": 1395
+    },
+    {
+      "epoch": 0.424348354738202,
+      "grad_norm": 1.88787841796875,
+      "learning_rate": 0.00015,
+      "loss": 4.223666667938232,
+      "step": 1396
+    },
+    {
+      "epoch": 0.4246523292043468,
+      "grad_norm": 1.6195083856582642,
+      "learning_rate": 0.00015,
+      "loss": 4.017670631408691,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4249563036704917,
+      "grad_norm": 1.8022451400756836,
+      "learning_rate": 0.00015,
+      "loss": 4.139994144439697,
+      "step": 1398
+    },
+    {
+      "epoch": 0.4252602781366365,
+      "grad_norm": 3.120497941970825,
+      "learning_rate": 0.00015,
+      "loss": 4.315942764282227,
+      "step": 1399
+    },
+    {
+      "epoch": 0.42556425260278136,
+      "grad_norm": 2.205777883529663,
+      "learning_rate": 0.00015,
+      "loss": 4.336596488952637,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4258682270689262,
+      "grad_norm": 2.1323351860046387,
+      "learning_rate": 0.00015,
+      "loss": 4.257810592651367,
+      "step": 1401
+    },
+    {
+      "epoch": 0.42617220153507107,
+      "grad_norm": 1.6776225566864014,
+      "learning_rate": 0.00015,
+      "loss": 4.169457912445068,
+      "step": 1402
+    },
+    {
+      "epoch": 0.4264761760012159,
+      "grad_norm": 2.2934231758117676,
+      "learning_rate": 0.00015,
+      "loss": 4.055765151977539,
+      "step": 1403
+    },
+    {
+      "epoch": 0.42678015046736073,
+      "grad_norm": 1.8692587614059448,
+      "learning_rate": 0.00015,
+      "loss": 4.068099021911621,
+      "step": 1404
+    },
+    {
+      "epoch": 0.42708412493350556,
+      "grad_norm": 1.6126461029052734,
+      "learning_rate": 0.00015,
+      "loss": 4.285520553588867,
+      "step": 1405
+    },
+    {
+      "epoch": 0.42738809939965045,
+      "grad_norm": 2.1855649948120117,
+      "learning_rate": 0.00015,
+      "loss": 4.112224578857422,
+      "step": 1406
+    },
+    {
+      "epoch": 0.4276920738657953,
+      "grad_norm": 1.9064407348632812,
+      "learning_rate": 0.00015,
+      "loss": 4.114408016204834,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4279960483319401,
+      "grad_norm": 2.1304142475128174,
+      "learning_rate": 0.00015,
+      "loss": 4.021907329559326,
+      "step": 1408
+    },
+    {
+      "epoch": 0.42830002279808493,
+      "grad_norm": 2.127925157546997,
+      "learning_rate": 0.00015,
+      "loss": 4.025974273681641,
+      "step": 1409
+    },
+    {
+      "epoch": 0.4286039972642298,
+      "grad_norm": 2.1613965034484863,
+      "learning_rate": 0.00015,
+      "loss": 3.7601377964019775,
+      "step": 1410
+    },
+    {
+      "epoch": 0.42890797173037465,
+      "grad_norm": 2.5019383430480957,
+      "learning_rate": 0.00015,
+      "loss": 4.083493232727051,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4292119461965195,
+      "grad_norm": 2.2067291736602783,
+      "learning_rate": 0.00015,
+      "loss": 4.327059745788574,
+      "step": 1412
+    },
+    {
+      "epoch": 0.42951592066266436,
+      "grad_norm": 1.8577791452407837,
+      "learning_rate": 0.00015,
+      "loss": 4.470882415771484,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4298198951288092,
+      "grad_norm": 4.022999286651611,
+      "learning_rate": 0.00015,
+      "loss": 4.580950736999512,
+      "step": 1414
+    },
+    {
+      "epoch": 0.430123869594954,
+      "grad_norm": 2.5104422569274902,
+      "learning_rate": 0.00015,
+      "loss": 4.080048561096191,
+      "step": 1415
+    },
+    {
+      "epoch": 0.43042784406109885,
+      "grad_norm": 2.014173746109009,
+      "learning_rate": 0.00015,
+      "loss": 4.100469589233398,
+      "step": 1416
+    },
+    {
+      "epoch": 0.43073181852724374,
+      "grad_norm": 2.613516330718994,
+      "learning_rate": 0.00015,
+      "loss": 3.8355319499969482,
+      "step": 1417
+    },
+    {
+      "epoch": 0.43103579299338857,
+      "grad_norm": 2.158290147781372,
+      "learning_rate": 0.00015,
+      "loss": 3.8920388221740723,
+      "step": 1418
+    },
+    {
+      "epoch": 0.4313397674595334,
+      "grad_norm": 2.2541112899780273,
+      "learning_rate": 0.00015,
+      "loss": 3.785411834716797,
+      "step": 1419
+    },
+    {
+      "epoch": 0.4316437419256782,
+      "grad_norm": 2.151209831237793,
+      "learning_rate": 0.00015,
+      "loss": 4.088793754577637,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4319477163918231,
+      "grad_norm": 1.9180150032043457,
+      "learning_rate": 0.00015,
+      "loss": 3.986478328704834,
+      "step": 1421
+    },
+    {
+      "epoch": 0.43225169085796794,
+      "grad_norm": 1.6735286712646484,
+      "learning_rate": 0.00015,
+      "loss": 4.191482067108154,
+      "step": 1422
+    },
+    {
+      "epoch": 0.43255566532411277,
+      "grad_norm": 1.905093789100647,
+      "learning_rate": 0.00015,
+      "loss": 4.081684589385986,
+      "step": 1423
+    },
+    {
+      "epoch": 0.4328596397902576,
+      "grad_norm": 1.8972699642181396,
+      "learning_rate": 0.00015,
+      "loss": 4.0188889503479,
+      "step": 1424
+    },
+    {
+      "epoch": 0.4331636142564025,
+      "grad_norm": 1.846854567527771,
+      "learning_rate": 0.00015,
+      "loss": 4.14583683013916,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4334675887225473,
+      "grad_norm": 2.017033576965332,
+      "learning_rate": 0.00015,
+      "loss": 4.233612537384033,
+      "step": 1426
+    },
+    {
+      "epoch": 0.43377156318869214,
+      "grad_norm": 2.2649686336517334,
+      "learning_rate": 0.00015,
+      "loss": 4.504868984222412,
+      "step": 1427
+    },
+    {
+      "epoch": 0.434075537654837,
+      "grad_norm": 1.7688179016113281,
+      "learning_rate": 0.00015,
+      "loss": 4.406624794006348,
+      "step": 1428
+    },
+    {
+      "epoch": 0.43437951212098186,
+      "grad_norm": 1.773761510848999,
+      "learning_rate": 0.00015,
+      "loss": 3.920217514038086,
+      "step": 1429
+    },
+    {
+      "epoch": 0.4346834865871267,
+      "grad_norm": 1.8407032489776611,
+      "learning_rate": 0.00015,
+      "loss": 4.233513832092285,
+      "step": 1430
+    },
+    {
+      "epoch": 0.4349874610532715,
+      "grad_norm": 3.0043537616729736,
+      "learning_rate": 0.00015,
+      "loss": 4.053088188171387,
+      "step": 1431
+    },
+    {
+      "epoch": 0.43529143551941635,
+      "grad_norm": 1.8753955364227295,
+      "learning_rate": 0.00015,
+      "loss": 4.1805009841918945,
+      "step": 1432
+    },
+    {
+      "epoch": 0.43559540998556123,
+      "grad_norm": 2.034454584121704,
+      "learning_rate": 0.00015,
+      "loss": 4.202044486999512,
+      "step": 1433
+    },
+    {
+      "epoch": 0.43589938445170606,
+      "grad_norm": 1.9841935634613037,
+      "learning_rate": 0.00015,
+      "loss": 4.1255784034729,
+      "step": 1434
+    },
+    {
+      "epoch": 0.4362033589178509,
+      "grad_norm": 1.696570634841919,
+      "learning_rate": 0.00015,
+      "loss": 4.116084098815918,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4365073333839957,
+      "grad_norm": 1.7869048118591309,
+      "learning_rate": 0.00015,
+      "loss": 3.8643229007720947,
+      "step": 1436
+    },
+    {
+      "epoch": 0.4368113078501406,
+      "grad_norm": 1.7332963943481445,
+      "learning_rate": 0.00015,
+      "loss": 4.160154342651367,
+      "step": 1437
+    },
+    {
+      "epoch": 0.43711528231628544,
+      "grad_norm": 1.7084729671478271,
+      "learning_rate": 0.00015,
+      "loss": 4.132291793823242,
+      "step": 1438
+    },
+    {
+      "epoch": 0.43741925678243027,
+      "grad_norm": 1.695223093032837,
+      "learning_rate": 0.00015,
+      "loss": 4.3258137702941895,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4377232312485751,
+      "grad_norm": 2.3917105197906494,
+      "learning_rate": 0.00015,
+      "loss": 4.068894386291504,
+      "step": 1440
+    },
+    {
+      "epoch": 0.43802720571472,
+      "grad_norm": 2.8022313117980957,
+      "learning_rate": 0.00015,
+      "loss": 4.506089210510254,
+      "step": 1441
+    },
+    {
+      "epoch": 0.4383311801808648,
+      "grad_norm": 1.7578727006912231,
+      "learning_rate": 0.00015,
+      "loss": 3.9265024662017822,
+      "step": 1442
+    },
+    {
+      "epoch": 0.43863515464700964,
+      "grad_norm": 2.149869918823242,
+      "learning_rate": 0.00015,
+      "loss": 4.046345233917236,
+      "step": 1443
+    },
+    {
+      "epoch": 0.43893912911315447,
+      "grad_norm": 2.0738227367401123,
+      "learning_rate": 0.00015,
+      "loss": 4.07380485534668,
+      "step": 1444
+    },
+    {
+      "epoch": 0.43924310357929935,
+      "grad_norm": 1.7357702255249023,
+      "learning_rate": 0.00015,
+      "loss": 4.111430644989014,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4395470780454442,
+      "grad_norm": 1.6537846326828003,
+      "learning_rate": 0.00015,
+      "loss": 4.0948686599731445,
+      "step": 1446
+    },
+    {
+      "epoch": 0.439851052511589,
+      "grad_norm": 1.9849975109100342,
+      "learning_rate": 0.00015,
+      "loss": 4.179037570953369,
+      "step": 1447
+    },
+    {
+      "epoch": 0.44015502697773384,
+      "grad_norm": 1.9715043306350708,
+      "learning_rate": 0.00015,
+      "loss": 3.9620285034179688,
+      "step": 1448
+    },
+    {
+      "epoch": 0.44045900144387873,
+      "grad_norm": 2.8953912258148193,
+      "learning_rate": 0.00015,
+      "loss": 4.3286848068237305,
+      "step": 1449
+    },
+    {
+      "epoch": 0.44076297591002356,
+      "grad_norm": 1.7863742113113403,
+      "learning_rate": 0.00015,
+      "loss": 4.100836753845215,
+      "step": 1450
+    },
+    {
+      "epoch": 0.4410669503761684,
+      "grad_norm": 1.901552438735962,
+      "learning_rate": 0.00015,
+      "loss": 4.182822227478027,
+      "step": 1451
+    },
+    {
+      "epoch": 0.4413709248423133,
+      "grad_norm": 2.3610923290252686,
+      "learning_rate": 0.00015,
+      "loss": 4.158679962158203,
+      "step": 1452
+    },
+    {
+      "epoch": 0.4416748993084581,
+      "grad_norm": 2.2818784713745117,
+      "learning_rate": 0.00015,
+      "loss": 4.105920791625977,
+      "step": 1453
+    },
+    {
+      "epoch": 0.44197887377460293,
+      "grad_norm": 1.584955096244812,
+      "learning_rate": 0.00015,
+      "loss": 4.240918159484863,
+      "step": 1454
+    },
+    {
+      "epoch": 0.44228284824074776,
+      "grad_norm": 1.92204749584198,
+      "learning_rate": 0.00015,
+      "loss": 4.199344158172607,
+      "step": 1455
+    },
+    {
+      "epoch": 0.44258682270689265,
+      "grad_norm": 2.0605437755584717,
+      "learning_rate": 0.00015,
+      "loss": 4.523133277893066,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4428907971730375,
+      "grad_norm": 2.198132276535034,
+      "learning_rate": 0.00015,
+      "loss": 4.131256580352783,
+      "step": 1457
+    },
+    {
+      "epoch": 0.4431947716391823,
+      "grad_norm": 2.001847982406616,
+      "learning_rate": 0.00015,
+      "loss": 3.9750583171844482,
+      "step": 1458
+    },
+    {
+      "epoch": 0.44349874610532714,
+      "grad_norm": 1.959446668624878,
+      "learning_rate": 0.00015,
+      "loss": 4.172065734863281,
+      "step": 1459
+    },
+    {
+      "epoch": 0.443802720571472,
+      "grad_norm": 1.8462470769882202,
+      "learning_rate": 0.00015,
+      "loss": 4.498308181762695,
+      "step": 1460
+    },
+    {
+      "epoch": 0.44410669503761685,
+      "grad_norm": 1.7062726020812988,
+      "learning_rate": 0.00015,
+      "loss": 4.099018573760986,
+      "step": 1461
+    },
+    {
+      "epoch": 0.4444106695037617,
+      "grad_norm": 2.699340581893921,
+      "learning_rate": 0.00015,
+      "loss": 4.241530418395996,
+      "step": 1462
+    },
+    {
+      "epoch": 0.4447146439699065,
+      "grad_norm": 2.2157435417175293,
+      "learning_rate": 0.00015,
+      "loss": 4.1595540046691895,
+      "step": 1463
+    },
+    {
+      "epoch": 0.4450186184360514,
+      "grad_norm": 1.6952439546585083,
+      "learning_rate": 0.00015,
+      "loss": 4.204911708831787,
+      "step": 1464
+    },
+    {
+      "epoch": 0.4453225929021962,
+      "grad_norm": 2.0300869941711426,
+      "learning_rate": 0.00015,
+      "loss": 4.072649955749512,
+      "step": 1465
+    },
+    {
+      "epoch": 0.44562656736834105,
+      "grad_norm": 2.151388645172119,
+      "learning_rate": 0.00015,
+      "loss": 4.36429500579834,
+      "step": 1466
+    },
+    {
+      "epoch": 0.4459305418344859,
+      "grad_norm": 2.2615368366241455,
+      "learning_rate": 0.00015,
+      "loss": 4.215755462646484,
+      "step": 1467
+    },
+    {
+      "epoch": 0.44623451630063077,
+      "grad_norm": 2.2193799018859863,
+      "learning_rate": 0.00015,
+      "loss": 4.112094879150391,
+      "step": 1468
+    },
+    {
+      "epoch": 0.4465384907667756,
+      "grad_norm": 1.9465091228485107,
+      "learning_rate": 0.00015,
+      "loss": 4.282754421234131,
+      "step": 1469
+    },
+    {
+      "epoch": 0.4468424652329204,
+      "grad_norm": 1.6252202987670898,
+      "learning_rate": 0.00015,
+      "loss": 4.152392387390137,
+      "step": 1470
+    },
+    {
+      "epoch": 0.44714643969906526,
+      "grad_norm": 2.2239181995391846,
+      "learning_rate": 0.00015,
+      "loss": 4.035510063171387,
+      "step": 1471
+    },
+    {
+      "epoch": 0.44745041416521014,
+      "grad_norm": 2.104738235473633,
+      "learning_rate": 0.00015,
+      "loss": 4.083586692810059,
+      "step": 1472
+    },
+    {
+      "epoch": 0.44775438863135497,
+      "grad_norm": 1.8182703256607056,
+      "learning_rate": 0.00015,
+      "loss": 4.27081298828125,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4480583630974998,
+      "grad_norm": 2.115696907043457,
+      "learning_rate": 0.00015,
+      "loss": 4.1502604484558105,
+      "step": 1474
+    },
+    {
+      "epoch": 0.44836233756364463,
+      "grad_norm": 1.6818941831588745,
+      "learning_rate": 0.00015,
+      "loss": 4.1901960372924805,
+      "step": 1475
+    },
+    {
+      "epoch": 0.4486663120297895,
+      "grad_norm": 1.959933876991272,
+      "learning_rate": 0.00015,
+      "loss": 4.293257713317871,
+      "step": 1476
+    },
+    {
+      "epoch": 0.44897028649593435,
+      "grad_norm": 1.518229365348816,
+      "learning_rate": 0.00015,
+      "loss": 4.241589546203613,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4492742609620792,
+      "grad_norm": 1.84169602394104,
+      "learning_rate": 0.00015,
+      "loss": 3.994833469390869,
+      "step": 1478
+    },
+    {
+      "epoch": 0.449578235428224,
+      "grad_norm": 1.753338098526001,
+      "learning_rate": 0.00015,
+      "loss": 3.967067241668701,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4498822098943689,
+      "grad_norm": 1.7331833839416504,
+      "learning_rate": 0.00015,
+      "loss": 4.2397074699401855,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4501861843605137,
+      "grad_norm": 3.0767722129821777,
+      "learning_rate": 0.00015,
+      "loss": 4.19664192199707,
+      "step": 1481
+    },
+    {
+      "epoch": 0.45049015882665855,
+      "grad_norm": 2.0414812564849854,
+      "learning_rate": 0.00015,
+      "loss": 4.1578569412231445,
+      "step": 1482
+    },
+    {
+      "epoch": 0.4507941332928034,
+      "grad_norm": 2.471980571746826,
+      "learning_rate": 0.00015,
+      "loss": 4.230874061584473,
+      "step": 1483
+    },
+    {
+      "epoch": 0.45109810775894826,
+      "grad_norm": 2.4603066444396973,
+      "learning_rate": 0.00015,
+      "loss": 3.8399596214294434,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4514020822250931,
+      "grad_norm": 1.8737504482269287,
+      "learning_rate": 0.00015,
+      "loss": 4.14688777923584,
+      "step": 1485
+    },
+    {
+      "epoch": 0.4517060566912379,
+      "grad_norm": 1.9794410467147827,
+      "learning_rate": 0.00015,
+      "loss": 4.170577049255371,
+      "step": 1486
+    },
+    {
+      "epoch": 0.45201003115738275,
+      "grad_norm": 1.9769762754440308,
+      "learning_rate": 0.00015,
+      "loss": 3.977240800857544,
+      "step": 1487
+    },
+    {
+      "epoch": 0.45231400562352764,
+      "grad_norm": 1.7836833000183105,
+      "learning_rate": 0.00015,
+      "loss": 4.161801338195801,
+      "step": 1488
+    },
+    {
+      "epoch": 0.45261798008967247,
+      "grad_norm": 1.9385381937026978,
+      "learning_rate": 0.00015,
+      "loss": 4.1907148361206055,
+      "step": 1489
+    },
+    {
+      "epoch": 0.4529219545558173,
+      "grad_norm": 3.427055597305298,
+      "learning_rate": 0.00015,
+      "loss": 4.605927467346191,
+      "step": 1490
+    },
+    {
+      "epoch": 0.4532259290219622,
+      "grad_norm": 2.337656259536743,
+      "learning_rate": 0.00015,
+      "loss": 4.166092872619629,
+      "step": 1491
+    },
+    {
+      "epoch": 0.453529903488107,
+      "grad_norm": 1.8724970817565918,
+      "learning_rate": 0.00015,
+      "loss": 4.241426944732666,
+      "step": 1492
+    },
+    {
+      "epoch": 0.45383387795425184,
+      "grad_norm": 2.180553436279297,
+      "learning_rate": 0.00015,
+      "loss": 4.05593204498291,
+      "step": 1493
+    },
+    {
+      "epoch": 0.45413785242039667,
+      "grad_norm": 1.9920824766159058,
+      "learning_rate": 0.00015,
+      "loss": 4.170358657836914,
+      "step": 1494
+    },
+    {
+      "epoch": 0.45444182688654156,
+      "grad_norm": 2.53837513923645,
+      "learning_rate": 0.00015,
+      "loss": 4.131795883178711,
+      "step": 1495
+    },
+    {
+      "epoch": 0.4547458013526864,
+      "grad_norm": 1.8016866445541382,
+      "learning_rate": 0.00015,
+      "loss": 4.321488857269287,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4550497758188312,
+      "grad_norm": 2.3698067665100098,
+      "learning_rate": 0.00015,
+      "loss": 4.303975582122803,
+      "step": 1497
+    },
+    {
+      "epoch": 0.45535375028497604,
+      "grad_norm": 1.8370994329452515,
+      "learning_rate": 0.00015,
+      "loss": 4.106391429901123,
+      "step": 1498
+    },
+    {
+      "epoch": 0.45565772475112093,
+      "grad_norm": 1.9031264781951904,
+      "learning_rate": 0.00015,
+      "loss": 3.9840316772460938,
+      "step": 1499
+    },
+    {
+      "epoch": 0.45596169921726576,
+      "grad_norm": 2.016136646270752,
+      "learning_rate": 0.00015,
+      "loss": 4.141274452209473,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4562656736834106,
+      "grad_norm": 2.6802477836608887,
+      "learning_rate": 0.00015,
+      "loss": 3.9779138565063477,
+      "step": 1501
+    },
+    {
+      "epoch": 0.4565696481495554,
+      "grad_norm": 1.771829605102539,
+      "learning_rate": 0.00015,
+      "loss": 4.164472579956055,
+      "step": 1502
+    },
+    {
+      "epoch": 0.4568736226157003,
+      "grad_norm": 1.6135309934616089,
+      "learning_rate": 0.00015,
+      "loss": 4.108955383300781,
+      "step": 1503
+    },
+    {
+      "epoch": 0.45717759708184513,
+      "grad_norm": 1.7662051916122437,
+      "learning_rate": 0.00015,
+      "loss": 4.333979606628418,
+      "step": 1504
+    },
+    {
+      "epoch": 0.45748157154798996,
+      "grad_norm": 1.8704208135604858,
+      "learning_rate": 0.00015,
+      "loss": 3.9974920749664307,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4577855460141348,
+      "grad_norm": 1.8002725839614868,
+      "learning_rate": 0.00015,
+      "loss": 3.9549174308776855,
+      "step": 1506
+    },
+    {
+      "epoch": 0.4580895204802797,
+      "grad_norm": 1.992700457572937,
+      "learning_rate": 0.00015,
+      "loss": 3.939211845397949,
+      "step": 1507
+    },
+    {
+      "epoch": 0.4583934949464245,
+      "grad_norm": 1.9790765047073364,
+      "learning_rate": 0.00015,
+      "loss": 4.100804805755615,
+      "step": 1508
+    },
+    {
+      "epoch": 0.45869746941256934,
+      "grad_norm": 1.6732966899871826,
+      "learning_rate": 0.00015,
+      "loss": 4.005144119262695,
+      "step": 1509
+    },
+    {
+      "epoch": 0.45900144387871417,
+      "grad_norm": 1.9715501070022583,
+      "learning_rate": 0.00015,
+      "loss": 4.114552021026611,
+      "step": 1510
+    },
+    {
+      "epoch": 0.45930541834485905,
+      "grad_norm": 1.9477407932281494,
+      "learning_rate": 0.00015,
+      "loss": 4.2671284675598145,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4596093928110039,
+      "grad_norm": 1.5729364156723022,
+      "learning_rate": 0.00015,
+      "loss": 4.03032922744751,
+      "step": 1512
+    },
+    {
+      "epoch": 0.4599133672771487,
+      "grad_norm": 4.073946475982666,
+      "learning_rate": 0.00015,
+      "loss": 4.322243690490723,
+      "step": 1513
+    },
+    {
+      "epoch": 0.46021734174329354,
+      "grad_norm": 1.9729983806610107,
+      "learning_rate": 0.00015,
+      "loss": 4.145951271057129,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4605213162094384,
+      "grad_norm": 1.7073591947555542,
+      "learning_rate": 0.00015,
+      "loss": 4.07159948348999,
+      "step": 1515
+    },
+    {
+      "epoch": 0.46082529067558325,
+      "grad_norm": 1.7636409997940063,
+      "learning_rate": 0.00015,
+      "loss": 4.161297798156738,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4611292651417281,
+      "grad_norm": 2.380549907684326,
+      "learning_rate": 0.00015,
+      "loss": 4.224286079406738,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4614332396078729,
+      "grad_norm": 1.9972162246704102,
+      "learning_rate": 0.00015,
+      "loss": 4.340700626373291,
+      "step": 1518
+    },
+    {
+      "epoch": 0.4617372140740178,
+      "grad_norm": 2.4851162433624268,
+      "learning_rate": 0.00015,
+      "loss": 3.990753173828125,
+      "step": 1519
+    },
+    {
+      "epoch": 0.46204118854016263,
+      "grad_norm": 1.7682558298110962,
+      "learning_rate": 0.00015,
+      "loss": 4.113886833190918,
+      "step": 1520
+    },
+    {
+      "epoch": 0.46234516300630746,
+      "grad_norm": 2.1126956939697266,
+      "learning_rate": 0.00015,
+      "loss": 4.06333065032959,
+      "step": 1521
+    },
+    {
+      "epoch": 0.4626491374724523,
+      "grad_norm": 1.6793979406356812,
+      "learning_rate": 0.00015,
+      "loss": 4.283230781555176,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4629531119385972,
+      "grad_norm": 2.1289913654327393,
+      "learning_rate": 0.00015,
+      "loss": 4.23714017868042,
+      "step": 1523
+    },
+    {
+      "epoch": 0.463257086404742,
+      "grad_norm": 2.1317129135131836,
+      "learning_rate": 0.00015,
+      "loss": 4.116940021514893,
+      "step": 1524
+    },
+    {
+      "epoch": 0.46356106087088683,
+      "grad_norm": 3.234846830368042,
+      "learning_rate": 0.00015,
+      "loss": 4.231685161590576,
+      "step": 1525
+    },
+    {
+      "epoch": 0.4638650353370317,
+      "grad_norm": 1.8869097232818604,
+      "learning_rate": 0.00015,
+      "loss": 3.9482460021972656,
+      "step": 1526
+    },
+    {
+      "epoch": 0.46416900980317655,
+      "grad_norm": 1.8360676765441895,
+      "learning_rate": 0.00015,
+      "loss": 4.356411457061768,
+      "step": 1527
+    },
+    {
+      "epoch": 0.4644729842693214,
+      "grad_norm": 1.6278525590896606,
+      "learning_rate": 0.00015,
+      "loss": 4.0491943359375,
+      "step": 1528
+    },
+    {
+      "epoch": 0.4647769587354662,
+      "grad_norm": 1.9716061353683472,
+      "learning_rate": 0.00015,
+      "loss": 4.276702880859375,
+      "step": 1529
+    },
+    {
+      "epoch": 0.4650809332016111,
+      "grad_norm": 2.1133062839508057,
+      "learning_rate": 0.00015,
+      "loss": 4.24643087387085,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4653849076677559,
+      "grad_norm": 1.8244991302490234,
+      "learning_rate": 0.00015,
+      "loss": 3.8125991821289062,
+      "step": 1531
+    },
+    {
+      "epoch": 0.46568888213390075,
+      "grad_norm": 1.6107333898544312,
+      "learning_rate": 0.00015,
+      "loss": 3.932375907897949,
+      "step": 1532
+    },
+    {
+      "epoch": 0.4659928566000456,
+      "grad_norm": 2.252056837081909,
+      "learning_rate": 0.00015,
+      "loss": 4.3183770179748535,
+      "step": 1533
+    },
+    {
+      "epoch": 0.46629683106619046,
+      "grad_norm": 2.0787315368652344,
+      "learning_rate": 0.00015,
+      "loss": 4.039055347442627,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4666008055323353,
+      "grad_norm": 1.848189115524292,
+      "learning_rate": 0.00015,
+      "loss": 4.554778099060059,
+      "step": 1535
+    },
+    {
+      "epoch": 0.4669047799984801,
+      "grad_norm": 2.271169424057007,
+      "learning_rate": 0.00015,
+      "loss": 3.925475597381592,
+      "step": 1536
+    },
+    {
+      "epoch": 0.46720875446462495,
+      "grad_norm": 1.7312628030776978,
+      "learning_rate": 0.00015,
+      "loss": 4.41420316696167,
+      "step": 1537
+    },
+    {
+      "epoch": 0.46751272893076984,
+      "grad_norm": 2.3496713638305664,
+      "learning_rate": 0.00015,
+      "loss": 4.236806869506836,
+      "step": 1538
+    },
+    {
+      "epoch": 0.46781670339691467,
+      "grad_norm": 2.308851480484009,
+      "learning_rate": 0.00015,
+      "loss": 4.394627094268799,
+      "step": 1539
+    },
+    {
+      "epoch": 0.4681206778630595,
+      "grad_norm": 1.8490279912948608,
+      "learning_rate": 0.00015,
+      "loss": 3.98945951461792,
+      "step": 1540
+    },
+    {
+      "epoch": 0.4684246523292043,
+      "grad_norm": 1.803397536277771,
+      "learning_rate": 0.00015,
+      "loss": 4.0500335693359375,
+      "step": 1541
+    },
+    {
+      "epoch": 0.4687286267953492,
+      "grad_norm": 2.2059991359710693,
+      "learning_rate": 0.00015,
+      "loss": 4.136713027954102,
+      "step": 1542
+    },
+    {
+      "epoch": 0.46903260126149404,
+      "grad_norm": 1.8286327123641968,
+      "learning_rate": 0.00015,
+      "loss": 3.9456710815429688,
+      "step": 1543
+    },
+    {
+      "epoch": 0.46933657572763887,
+      "grad_norm": 1.5285968780517578,
+      "learning_rate": 0.00015,
+      "loss": 4.718966484069824,
+      "step": 1544
+    },
+    {
+      "epoch": 0.4696405501937837,
+      "grad_norm": 2.402273178100586,
+      "learning_rate": 0.00015,
+      "loss": 4.2217254638671875,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4699445246599286,
+      "grad_norm": 1.7056288719177246,
+      "learning_rate": 0.00015,
+      "loss": 3.9872419834136963,
+      "step": 1546
+    },
+    {
+      "epoch": 0.4702484991260734,
+      "grad_norm": 1.9320281744003296,
+      "learning_rate": 0.00015,
+      "loss": 3.989673614501953,
+      "step": 1547
+    },
+    {
+      "epoch": 0.47055247359221825,
+      "grad_norm": 3.150632381439209,
+      "learning_rate": 0.00015,
+      "loss": 4.0228800773620605,
+      "step": 1548
+    },
+    {
+      "epoch": 0.4708564480583631,
+      "grad_norm": 1.7547643184661865,
+      "learning_rate": 0.00015,
+      "loss": 4.244220733642578,
+      "step": 1549
+    },
+    {
+      "epoch": 0.47116042252450796,
+      "grad_norm": 3.365715980529785,
+      "learning_rate": 0.00015,
+      "loss": 4.181885719299316,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4714643969906528,
+      "grad_norm": 1.6299647092819214,
+      "learning_rate": 0.00015,
+      "loss": 4.077173233032227,
+      "step": 1551
+    },
+    {
+      "epoch": 0.4717683714567976,
+      "grad_norm": 2.083721160888672,
+      "learning_rate": 0.00015,
+      "loss": 4.044068336486816,
+      "step": 1552
+    },
+    {
+      "epoch": 0.47207234592294245,
+      "grad_norm": 2.4484665393829346,
+      "learning_rate": 0.00015,
+      "loss": 4.1517486572265625,
+      "step": 1553
+    },
+    {
+      "epoch": 0.47237632038908733,
+      "grad_norm": 2.1219048500061035,
+      "learning_rate": 0.00015,
+      "loss": 4.131948947906494,
+      "step": 1554
+    },
+    {
+      "epoch": 0.47268029485523216,
+      "grad_norm": 2.0985467433929443,
+      "learning_rate": 0.00015,
+      "loss": 3.8195886611938477,
+      "step": 1555
+    },
+    {
+      "epoch": 0.472984269321377,
+      "grad_norm": 1.9423420429229736,
+      "learning_rate": 0.00015,
+      "loss": 4.11447811126709,
+      "step": 1556
+    },
+    {
+      "epoch": 0.4732882437875218,
+      "grad_norm": 1.643110752105713,
+      "learning_rate": 0.00015,
+      "loss": 4.146320819854736,
+      "step": 1557
+    },
+    {
+      "epoch": 0.4735922182536667,
+      "grad_norm": 1.616041660308838,
+      "learning_rate": 0.00015,
+      "loss": 4.253957748413086,
+      "step": 1558
+    },
+    {
+      "epoch": 0.47389619271981154,
+      "grad_norm": 1.8653925657272339,
+      "learning_rate": 0.00015,
+      "loss": 3.9709935188293457,
+      "step": 1559
+    },
+    {
+      "epoch": 0.47420016718595637,
+      "grad_norm": 1.8392813205718994,
+      "learning_rate": 0.00015,
+      "loss": 3.9680306911468506,
+      "step": 1560
+    },
+    {
+      "epoch": 0.4745041416521012,
+      "grad_norm": 1.8709595203399658,
+      "learning_rate": 0.00015,
+      "loss": 4.324101448059082,
+      "step": 1561
+    },
+    {
+      "epoch": 0.4748081161182461,
+      "grad_norm": 2.2005176544189453,
+      "learning_rate": 0.00015,
+      "loss": 4.266378879547119,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4751120905843909,
+      "grad_norm": 1.709680199623108,
+      "learning_rate": 0.00015,
+      "loss": 4.078587532043457,
+      "step": 1563
+    },
+    {
+      "epoch": 0.47541606505053574,
+      "grad_norm": 1.8984386920928955,
+      "learning_rate": 0.00015,
+      "loss": 3.9547741413116455,
+      "step": 1564
+    },
+    {
+      "epoch": 0.4757200395166806,
+      "grad_norm": 1.8166203498840332,
+      "learning_rate": 0.00015,
+      "loss": 3.939985752105713,
+      "step": 1565
+    },
+    {
+      "epoch": 0.47602401398282546,
+      "grad_norm": 1.8323556184768677,
+      "learning_rate": 0.00015,
+      "loss": 3.918497085571289,
+      "step": 1566
+    },
+    {
+      "epoch": 0.4763279884489703,
+      "grad_norm": 2.120941638946533,
+      "learning_rate": 0.00015,
+      "loss": 4.267053127288818,
+      "step": 1567
+    },
+    {
+      "epoch": 0.4766319629151151,
+      "grad_norm": 1.9001502990722656,
+      "learning_rate": 0.00015,
+      "loss": 4.230142593383789,
+      "step": 1568
+    },
+    {
+      "epoch": 0.47693593738126,
+      "grad_norm": 1.910078525543213,
+      "learning_rate": 0.00015,
+      "loss": 4.290778160095215,
+      "step": 1569
+    },
+    {
+      "epoch": 0.47723991184740483,
+      "grad_norm": 1.758470058441162,
+      "learning_rate": 0.00015,
+      "loss": 4.212425231933594,
+      "step": 1570
+    },
+    {
+      "epoch": 0.47754388631354966,
+      "grad_norm": 1.6282622814178467,
+      "learning_rate": 0.00015,
+      "loss": 4.223325729370117,
+      "step": 1571
+    },
+    {
+      "epoch": 0.4778478607796945,
+      "grad_norm": 2.749488592147827,
+      "learning_rate": 0.00015,
+      "loss": 3.9993157386779785,
+      "step": 1572
+    },
+    {
+      "epoch": 0.4781518352458394,
+      "grad_norm": 2.0481839179992676,
+      "learning_rate": 0.00015,
+      "loss": 4.1050519943237305,
+      "step": 1573
+    },
+    {
+      "epoch": 0.4784558097119842,
+      "grad_norm": 2.327662706375122,
+      "learning_rate": 0.00015,
+      "loss": 4.190561294555664,
+      "step": 1574
+    },
+    {
+      "epoch": 0.47875978417812903,
+      "grad_norm": 2.45314884185791,
+      "learning_rate": 0.00015,
+      "loss": 4.128019332885742,
+      "step": 1575
+    },
+    {
+      "epoch": 0.47906375864427386,
+      "grad_norm": 1.8194535970687866,
+      "learning_rate": 0.00015,
+      "loss": 4.3149919509887695,
+      "step": 1576
+    },
+    {
+      "epoch": 0.47936773311041875,
+      "grad_norm": 1.7792165279388428,
+      "learning_rate": 0.00015,
+      "loss": 4.177857398986816,
+      "step": 1577
+    },
+    {
+      "epoch": 0.4796717075765636,
+      "grad_norm": 2.241546630859375,
+      "learning_rate": 0.00015,
+      "loss": 4.288545608520508,
+      "step": 1578
+    },
+    {
+      "epoch": 0.4799756820427084,
+      "grad_norm": 3.8725967407226562,
+      "learning_rate": 0.00015,
+      "loss": 4.119723320007324,
+      "step": 1579
+    },
+    {
+      "epoch": 0.48027965650885324,
+      "grad_norm": 2.4643237590789795,
+      "learning_rate": 0.00015,
+      "loss": 4.289543628692627,
+      "step": 1580
+    },
+    {
+      "epoch": 0.4805836309749981,
+      "grad_norm": 2.92522931098938,
+      "learning_rate": 0.00015,
+      "loss": 4.157250881195068,
+      "step": 1581
+    },
+    {
+      "epoch": 0.48088760544114295,
+      "grad_norm": 2.6659297943115234,
+      "learning_rate": 0.00015,
+      "loss": 4.050342082977295,
+      "step": 1582
+    },
+    {
+      "epoch": 0.4811915799072878,
+      "grad_norm": 1.9758719205856323,
+      "learning_rate": 0.00015,
+      "loss": 3.9531397819519043,
+      "step": 1583
+    },
+    {
+      "epoch": 0.4814955543734326,
+      "grad_norm": 2.242856025695801,
+      "learning_rate": 0.00015,
+      "loss": 4.0514726638793945,
+      "step": 1584
+    },
+    {
+      "epoch": 0.4817995288395775,
+      "grad_norm": 1.4585992097854614,
+      "learning_rate": 0.00015,
+      "loss": 4.096534729003906,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4821035033057223,
+      "grad_norm": 1.936619520187378,
+      "learning_rate": 0.00015,
+      "loss": 4.20575475692749,
+      "step": 1586
+    },
+    {
+      "epoch": 0.48240747777186715,
+      "grad_norm": 1.985442876815796,
+      "learning_rate": 0.00015,
+      "loss": 3.964066505432129,
+      "step": 1587
+    },
+    {
+      "epoch": 0.482711452238012,
+      "grad_norm": 1.7622941732406616,
+      "learning_rate": 0.00015,
+      "loss": 4.251713275909424,
+      "step": 1588
+    },
+    {
+      "epoch": 0.48301542670415687,
+      "grad_norm": 2.039592981338501,
+      "learning_rate": 0.00015,
+      "loss": 4.060670852661133,
+      "step": 1589
+    },
+    {
+      "epoch": 0.4833194011703017,
+      "grad_norm": 2.392700672149658,
+      "learning_rate": 0.00015,
+      "loss": 4.111145496368408,
+      "step": 1590
+    },
+    {
+      "epoch": 0.48362337563644653,
+      "grad_norm": 1.6592389345169067,
+      "learning_rate": 0.00015,
+      "loss": 4.184656143188477,
+      "step": 1591
+    },
+    {
+      "epoch": 0.48392735010259136,
+      "grad_norm": 2.7381324768066406,
+      "learning_rate": 0.00015,
+      "loss": 3.937757968902588,
+      "step": 1592
+    },
+    {
+      "epoch": 0.48423132456873624,
+      "grad_norm": 2.05531907081604,
+      "learning_rate": 0.00015,
+      "loss": 4.214405059814453,
+      "step": 1593
+    },
+    {
+      "epoch": 0.4845352990348811,
+      "grad_norm": 1.8974171876907349,
+      "learning_rate": 0.00015,
+      "loss": 4.435897350311279,
+      "step": 1594
+    },
+    {
+      "epoch": 0.4848392735010259,
+      "grad_norm": 2.1932101249694824,
+      "learning_rate": 0.00015,
+      "loss": 4.448761463165283,
+      "step": 1595
+    },
+    {
+      "epoch": 0.48514324796717073,
+      "grad_norm": 2.3539421558380127,
+      "learning_rate": 0.00015,
+      "loss": 4.118188381195068,
+      "step": 1596
+    },
+    {
+      "epoch": 0.4854472224333156,
+      "grad_norm": 2.006777286529541,
+      "learning_rate": 0.00015,
+      "loss": 3.844573974609375,
+      "step": 1597
+    },
+    {
+      "epoch": 0.48575119689946045,
+      "grad_norm": 1.6563068628311157,
+      "learning_rate": 0.00015,
+      "loss": 4.368903636932373,
+      "step": 1598
+    },
+    {
+      "epoch": 0.4860551713656053,
+      "grad_norm": 2.013725996017456,
+      "learning_rate": 0.00015,
+      "loss": 4.000759124755859,
+      "step": 1599
+    },
+    {
+      "epoch": 0.4863591458317501,
+      "grad_norm": 2.296527624130249,
+      "learning_rate": 0.00015,
+      "loss": 3.8645148277282715,
+      "step": 1600
+    },
+    {
+      "epoch": 0.486663120297895,
+      "grad_norm": 2.05281662940979,
+      "learning_rate": 0.00015,
+      "loss": 4.287120819091797,
+      "step": 1601
+    },
+    {
+      "epoch": 0.4869670947640398,
+      "grad_norm": 1.8289650678634644,
+      "learning_rate": 0.00015,
+      "loss": 4.184939384460449,
+      "step": 1602
+    },
+    {
+      "epoch": 0.48727106923018465,
+      "grad_norm": 2.0394890308380127,
+      "learning_rate": 0.00015,
+      "loss": 4.174135208129883,
+      "step": 1603
+    },
+    {
+      "epoch": 0.48757504369632954,
+      "grad_norm": 2.1272189617156982,
+      "learning_rate": 0.00015,
+      "loss": 4.033539295196533,
+      "step": 1604
+    },
+    {
+      "epoch": 0.48787901816247436,
+      "grad_norm": 1.8586506843566895,
+      "learning_rate": 0.00015,
+      "loss": 4.0789947509765625,
+      "step": 1605
+    },
+    {
+      "epoch": 0.4881829926286192,
+      "grad_norm": 1.6590770483016968,
+      "learning_rate": 0.00015,
+      "loss": 4.007596492767334,
+      "step": 1606
+    },
+    {
+      "epoch": 0.488486967094764,
+      "grad_norm": 2.313508987426758,
+      "learning_rate": 0.00015,
+      "loss": 4.09402322769165,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4887909415609089,
+      "grad_norm": 1.8046866655349731,
+      "learning_rate": 0.00015,
+      "loss": 4.11134147644043,
+      "step": 1608
+    },
+    {
+      "epoch": 0.48909491602705374,
+      "grad_norm": 1.9055800437927246,
+      "learning_rate": 0.00015,
+      "loss": 3.8357815742492676,
+      "step": 1609
+    },
+    {
+      "epoch": 0.48939889049319857,
+      "grad_norm": 2.045945882797241,
+      "learning_rate": 0.00015,
+      "loss": 4.35378360748291,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4897028649593434,
+      "grad_norm": 1.7036305665969849,
+      "learning_rate": 0.00015,
+      "loss": 4.254256248474121,
+      "step": 1611
+    },
+    {
+      "epoch": 0.4900068394254883,
+      "grad_norm": 1.6609578132629395,
+      "learning_rate": 0.00015,
+      "loss": 4.139028072357178,
+      "step": 1612
+    },
+    {
+      "epoch": 0.4903108138916331,
+      "grad_norm": 1.958155870437622,
+      "learning_rate": 0.00015,
+      "loss": 3.9544296264648438,
+      "step": 1613
+    },
+    {
+      "epoch": 0.49061478835777794,
+      "grad_norm": 1.713308334350586,
+      "learning_rate": 0.00015,
+      "loss": 4.070626735687256,
+      "step": 1614
+    },
+    {
+      "epoch": 0.49091876282392277,
+      "grad_norm": 1.623005747795105,
+      "learning_rate": 0.00015,
+      "loss": 4.306579113006592,
+      "step": 1615
+    },
+    {
+      "epoch": 0.49122273729006766,
+      "grad_norm": 1.9919580221176147,
+      "learning_rate": 0.00015,
+      "loss": 4.184858798980713,
+      "step": 1616
+    },
+    {
+      "epoch": 0.4915267117562125,
+      "grad_norm": 2.0527443885803223,
+      "learning_rate": 0.00015,
+      "loss": 4.3444976806640625,
+      "step": 1617
+    },
+    {
+      "epoch": 0.4918306862223573,
+      "grad_norm": 2.432332754135132,
+      "learning_rate": 0.00015,
+      "loss": 4.019991874694824,
+      "step": 1618
+    },
+    {
+      "epoch": 0.49213466068850215,
+      "grad_norm": 1.891347050666809,
+      "learning_rate": 0.00015,
+      "loss": 4.116759300231934,
+      "step": 1619
+    },
+    {
+      "epoch": 0.49243863515464703,
+      "grad_norm": 1.7043830156326294,
+      "learning_rate": 0.00015,
+      "loss": 3.9404244422912598,
+      "step": 1620
+    },
+    {
+      "epoch": 0.49274260962079186,
+      "grad_norm": 2.1888720989227295,
+      "learning_rate": 0.00015,
+      "loss": 4.290994167327881,
+      "step": 1621
+    },
+    {
+      "epoch": 0.4930465840869367,
+      "grad_norm": 1.9937621355056763,
+      "learning_rate": 0.00015,
+      "loss": 4.474332809448242,
+      "step": 1622
+    },
+    {
+      "epoch": 0.4933505585530815,
+      "grad_norm": 2.407257318496704,
+      "learning_rate": 0.00015,
+      "loss": 4.030854225158691,
+      "step": 1623
+    },
+    {
+      "epoch": 0.4936545330192264,
+      "grad_norm": 1.7093148231506348,
+      "learning_rate": 0.00015,
+      "loss": 4.008371353149414,
+      "step": 1624
+    },
+    {
+      "epoch": 0.49395850748537123,
+      "grad_norm": 1.9812960624694824,
+      "learning_rate": 0.00015,
+      "loss": 4.135840892791748,
+      "step": 1625
+    },
+    {
+      "epoch": 0.49426248195151606,
+      "grad_norm": 1.8236570358276367,
+      "learning_rate": 0.00015,
+      "loss": 3.953756332397461,
+      "step": 1626
+    },
+    {
+      "epoch": 0.4945664564176609,
+      "grad_norm": 1.9534626007080078,
+      "learning_rate": 0.00015,
+      "loss": 3.6903862953186035,
+      "step": 1627
+    },
+    {
+      "epoch": 0.4948704308838058,
+      "grad_norm": 1.7371422052383423,
+      "learning_rate": 0.00015,
+      "loss": 4.146600246429443,
+      "step": 1628
+    },
+    {
+      "epoch": 0.4951744053499506,
+      "grad_norm": 1.8942281007766724,
+      "learning_rate": 0.00015,
+      "loss": 4.171694755554199,
+      "step": 1629
+    },
+    {
+      "epoch": 0.49547837981609544,
+      "grad_norm": 1.7443628311157227,
+      "learning_rate": 0.00015,
+      "loss": 4.103346824645996,
+      "step": 1630
+    },
+    {
+      "epoch": 0.49578235428224027,
+      "grad_norm": 1.9230849742889404,
+      "learning_rate": 0.00015,
+      "loss": 3.82694149017334,
+      "step": 1631
+    },
+    {
+      "epoch": 0.49608632874838515,
+      "grad_norm": 1.724745512008667,
+      "learning_rate": 0.00015,
+      "loss": 4.1363396644592285,
+      "step": 1632
+    },
+    {
+      "epoch": 0.49639030321453,
+      "grad_norm": 1.688772201538086,
+      "learning_rate": 0.00015,
+      "loss": 3.9784562587738037,
+      "step": 1633
+    },
+    {
+      "epoch": 0.4966942776806748,
+      "grad_norm": 1.8235375881195068,
+      "learning_rate": 0.00015,
+      "loss": 3.929655075073242,
+      "step": 1634
+    },
+    {
+      "epoch": 0.49699825214681964,
+      "grad_norm": 1.6541582345962524,
+      "learning_rate": 0.00015,
+      "loss": 4.049057960510254,
+      "step": 1635
+    },
+    {
+      "epoch": 0.4973022266129645,
+      "grad_norm": 2.0990023612976074,
+      "learning_rate": 0.00015,
+      "loss": 3.9777731895446777,
+      "step": 1636
+    },
+    {
+      "epoch": 0.49760620107910936,
+      "grad_norm": 1.7404930591583252,
+      "learning_rate": 0.00015,
+      "loss": 4.11392879486084,
+      "step": 1637
+    },
+    {
+      "epoch": 0.4979101755452542,
+      "grad_norm": 1.6381345987319946,
+      "learning_rate": 0.00015,
+      "loss": 4.120820999145508,
+      "step": 1638
+    },
+    {
+      "epoch": 0.498214150011399,
+      "grad_norm": 1.6543736457824707,
+      "learning_rate": 0.00015,
+      "loss": 3.924229621887207,
+      "step": 1639
+    },
+    {
+      "epoch": 0.4985181244775439,
+      "grad_norm": 1.9583319425582886,
+      "learning_rate": 0.00015,
+      "loss": 3.9547853469848633,
+      "step": 1640
+    },
+    {
+      "epoch": 0.49882209894368873,
+      "grad_norm": 2.024052381515503,
+      "learning_rate": 0.00015,
+      "loss": 4.395962715148926,
+      "step": 1641
+    },
+    {
+      "epoch": 0.49912607340983356,
+      "grad_norm": 1.7071644067764282,
+      "learning_rate": 0.00015,
+      "loss": 4.219705581665039,
+      "step": 1642
+    },
+    {
+      "epoch": 0.49943004787597844,
+      "grad_norm": 2.2150745391845703,
+      "learning_rate": 0.00015,
+      "loss": 4.30179500579834,
+      "step": 1643
+    },
+    {
+      "epoch": 0.4997340223421233,
+      "grad_norm": 1.6711382865905762,
+      "learning_rate": 0.00015,
+      "loss": 4.2185378074646,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5000379968082681,
+      "grad_norm": 2.0096445083618164,
+      "learning_rate": 0.00015,
+      "loss": 4.201049327850342,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5003419712744129,
+      "grad_norm": 2.0067570209503174,
+      "learning_rate": 0.00015,
+      "loss": 4.234549045562744,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5006459457405578,
+      "grad_norm": 1.777790904045105,
+      "learning_rate": 0.00015,
+      "loss": 3.9791746139526367,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5009499202067026,
+      "grad_norm": 1.9716025590896606,
+      "learning_rate": 0.00015,
+      "loss": 4.143360137939453,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5012538946728475,
+      "grad_norm": 1.807313084602356,
+      "learning_rate": 0.00015,
+      "loss": 4.021626949310303,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5015578691389924,
+      "grad_norm": 1.8576842546463013,
+      "learning_rate": 0.00015,
+      "loss": 4.297464370727539,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5018618436051372,
+      "grad_norm": 2.0815391540527344,
+      "learning_rate": 0.00015,
+      "loss": 4.1967244148254395,
+      "step": 1651
+    },
+    {
+      "epoch": 0.502165818071282,
+      "grad_norm": 1.9786092042922974,
+      "learning_rate": 0.00015,
+      "loss": 4.230411529541016,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5024697925374269,
+      "grad_norm": 1.947917103767395,
+      "learning_rate": 0.00015,
+      "loss": 4.143929481506348,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5027737670035717,
+      "grad_norm": 2.0179405212402344,
+      "learning_rate": 0.00015,
+      "loss": 4.063713550567627,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5030777414697165,
+      "grad_norm": 3.428227186203003,
+      "learning_rate": 0.00015,
+      "loss": 4.208183288574219,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5033817159358613,
+      "grad_norm": 1.7879821062088013,
+      "learning_rate": 0.00015,
+      "loss": 3.8356127738952637,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5036856904020063,
+      "grad_norm": 1.6827938556671143,
+      "learning_rate": 0.00015,
+      "loss": 3.6757450103759766,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5039896648681511,
+      "grad_norm": 1.5272655487060547,
+      "learning_rate": 0.00015,
+      "loss": 3.869305372238159,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5042936393342959,
+      "grad_norm": 2.113154888153076,
+      "learning_rate": 0.00015,
+      "loss": 4.114598274230957,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5045976138004408,
+      "grad_norm": 2.056516170501709,
+      "learning_rate": 0.00015,
+      "loss": 4.074169635772705,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5049015882665856,
+      "grad_norm": 2.6815690994262695,
+      "learning_rate": 0.00015,
+      "loss": 4.170917510986328,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5052055627327304,
+      "grad_norm": 2.1671226024627686,
+      "learning_rate": 0.00015,
+      "loss": 4.218400955200195,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5055095371988753,
+      "grad_norm": 2.025648355484009,
+      "learning_rate": 0.00015,
+      "loss": 4.037323951721191,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5058135116650201,
+      "grad_norm": 2.0837721824645996,
+      "learning_rate": 0.00015,
+      "loss": 3.7895689010620117,
+      "step": 1664
+    },
+    {
+      "epoch": 0.506117486131165,
+      "grad_norm": 2.746493339538574,
+      "learning_rate": 0.00015,
+      "loss": 3.929539918899536,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5064214605973099,
+      "grad_norm": 2.134633779525757,
+      "learning_rate": 0.00015,
+      "loss": 3.947176933288574,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5067254350634547,
+      "grad_norm": 1.9383728504180908,
+      "learning_rate": 0.00015,
+      "loss": 4.200277328491211,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5070294095295995,
+      "grad_norm": 1.9043116569519043,
+      "learning_rate": 0.00015,
+      "loss": 4.2036919593811035,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5073333839957443,
+      "grad_norm": 1.676783561706543,
+      "learning_rate": 0.00015,
+      "loss": 4.124678611755371,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5076373584618892,
+      "grad_norm": 2.1080470085144043,
+      "learning_rate": 0.00015,
+      "loss": 3.8944921493530273,
+      "step": 1670
+    },
+    {
+      "epoch": 0.507941332928034,
+      "grad_norm": 2.0180280208587646,
+      "learning_rate": 0.00015,
+      "loss": 4.095719814300537,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5082453073941788,
+      "grad_norm": 1.7777819633483887,
+      "learning_rate": 0.00015,
+      "loss": 4.401481628417969,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5085492818603238,
+      "grad_norm": 2.0122227668762207,
+      "learning_rate": 0.00015,
+      "loss": 3.940009117126465,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5088532563264686,
+      "grad_norm": 2.0286800861358643,
+      "learning_rate": 0.00015,
+      "loss": 3.9894919395446777,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5091572307926134,
+      "grad_norm": 1.6453368663787842,
+      "learning_rate": 0.00015,
+      "loss": 4.119282245635986,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5094612052587583,
+      "grad_norm": 2.445148229598999,
+      "learning_rate": 0.00015,
+      "loss": 3.9753823280334473,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5097651797249031,
+      "grad_norm": 2.5450193881988525,
+      "learning_rate": 0.00015,
+      "loss": 4.1424479484558105,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5100691541910479,
+      "grad_norm": 1.872321605682373,
+      "learning_rate": 0.00015,
+      "loss": 3.821399450302124,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5103731286571928,
+      "grad_norm": 1.8487789630889893,
+      "learning_rate": 0.00015,
+      "loss": 4.007262229919434,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5106771031233377,
+      "grad_norm": 1.9029178619384766,
+      "learning_rate": 0.00015,
+      "loss": 4.087822914123535,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5109810775894825,
+      "grad_norm": 1.9204846620559692,
+      "learning_rate": 0.00015,
+      "loss": 3.9775519371032715,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5112850520556274,
+      "grad_norm": 2.0020978450775146,
+      "learning_rate": 0.00015,
+      "loss": 4.079058647155762,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5115890265217722,
+      "grad_norm": 2.0765879154205322,
+      "learning_rate": 0.00015,
+      "loss": 3.8106396198272705,
+      "step": 1683
+    },
+    {
+      "epoch": 0.511893000987917,
+      "grad_norm": 2.5066418647766113,
+      "learning_rate": 0.00015,
+      "loss": 3.7626888751983643,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5121969754540618,
+      "grad_norm": 2.264890670776367,
+      "learning_rate": 0.00015,
+      "loss": 4.257828712463379,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5125009499202067,
+      "grad_norm": 1.8571083545684814,
+      "learning_rate": 0.00015,
+      "loss": 4.025185585021973,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5128049243863515,
+      "grad_norm": 1.9406288862228394,
+      "learning_rate": 0.00015,
+      "loss": 3.913438320159912,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5131088988524964,
+      "grad_norm": 2.283121109008789,
+      "learning_rate": 0.00015,
+      "loss": 4.217222213745117,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5134128733186413,
+      "grad_norm": 1.9238258600234985,
+      "learning_rate": 0.00015,
+      "loss": 4.086227893829346,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5137168477847861,
+      "grad_norm": 1.9317116737365723,
+      "learning_rate": 0.00015,
+      "loss": 4.041914463043213,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5140208222509309,
+      "grad_norm": 1.674665927886963,
+      "learning_rate": 0.00015,
+      "loss": 3.9255332946777344,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5143247967170758,
+      "grad_norm": 1.7832942008972168,
+      "learning_rate": 0.00015,
+      "loss": 4.352984428405762,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5146287711832206,
+      "grad_norm": 2.5971779823303223,
+      "learning_rate": 0.00015,
+      "loss": 4.096341133117676,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5149327456493654,
+      "grad_norm": 4.793689250946045,
+      "learning_rate": 0.00015,
+      "loss": 3.8930444717407227,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5152367201155102,
+      "grad_norm": 1.7985873222351074,
+      "learning_rate": 0.00015,
+      "loss": 4.108163833618164,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5155406945816552,
+      "grad_norm": 2.0064985752105713,
+      "learning_rate": 0.00015,
+      "loss": 3.8716063499450684,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5158446690478,
+      "grad_norm": 2.1424038410186768,
+      "learning_rate": 0.00015,
+      "loss": 4.069830894470215,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5161486435139448,
+      "grad_norm": 1.9811183214187622,
+      "learning_rate": 0.00015,
+      "loss": 3.7200429439544678,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5164526179800897,
+      "grad_norm": 1.8026741743087769,
+      "learning_rate": 0.00015,
+      "loss": 3.8046317100524902,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5167565924462345,
+      "grad_norm": 2.2181777954101562,
+      "learning_rate": 0.00015,
+      "loss": 4.276614665985107,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5170605669123793,
+      "grad_norm": 2.254011392593384,
+      "learning_rate": 0.00015,
+      "loss": 4.017321586608887,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5173645413785242,
+      "grad_norm": 2.005218505859375,
+      "learning_rate": 0.00015,
+      "loss": 4.193414688110352,
+      "step": 1702
+    },
+    {
+      "epoch": 0.517668515844669,
+      "grad_norm": 1.6464335918426514,
+      "learning_rate": 0.00015,
+      "loss": 4.252065181732178,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5179724903108139,
+      "grad_norm": 2.066028118133545,
+      "learning_rate": 0.00015,
+      "loss": 4.087015151977539,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5182764647769588,
+      "grad_norm": 2.099611759185791,
+      "learning_rate": 0.00015,
+      "loss": 4.141718864440918,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5185804392431036,
+      "grad_norm": 2.059425115585327,
+      "learning_rate": 0.00015,
+      "loss": 4.072841167449951,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5188844137092484,
+      "grad_norm": 2.4786951541900635,
+      "learning_rate": 0.00015,
+      "loss": 4.150832176208496,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5191883881753933,
+      "grad_norm": 2.5313868522644043,
+      "learning_rate": 0.00015,
+      "loss": 3.8542072772979736,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5194923626415381,
+      "grad_norm": 2.209953546524048,
+      "learning_rate": 0.00015,
+      "loss": 3.8910024166107178,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5197963371076829,
+      "grad_norm": 2.5466701984405518,
+      "learning_rate": 0.00015,
+      "loss": 3.7541770935058594,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5201003115738277,
+      "grad_norm": 1.8166781663894653,
+      "learning_rate": 0.00015,
+      "loss": 4.341780662536621,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5204042860399727,
+      "grad_norm": 2.167344808578491,
+      "learning_rate": 0.00015,
+      "loss": 3.747084617614746,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5207082605061175,
+      "grad_norm": 1.874870777130127,
+      "learning_rate": 0.00015,
+      "loss": 3.8597774505615234,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5210122349722623,
+      "grad_norm": 2.013676881790161,
+      "learning_rate": 0.00015,
+      "loss": 4.002860069274902,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5213162094384072,
+      "grad_norm": 1.8548755645751953,
+      "learning_rate": 0.00015,
+      "loss": 4.111102104187012,
+      "step": 1715
+    },
+    {
+      "epoch": 0.521620183904552,
+      "grad_norm": 1.8581019639968872,
+      "learning_rate": 0.00015,
+      "loss": 4.060452461242676,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5219241583706968,
+      "grad_norm": 2.7270286083221436,
+      "learning_rate": 0.00015,
+      "loss": 4.3279218673706055,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5222281328368417,
+      "grad_norm": 2.1587228775024414,
+      "learning_rate": 0.00015,
+      "loss": 4.329817295074463,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5225321073029866,
+      "grad_norm": 1.8935465812683105,
+      "learning_rate": 0.00015,
+      "loss": 4.374145984649658,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5228360817691314,
+      "grad_norm": 2.184800386428833,
+      "learning_rate": 0.00015,
+      "loss": 4.085572242736816,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5231400562352763,
+      "grad_norm": 2.215609073638916,
+      "learning_rate": 0.00015,
+      "loss": 4.116372585296631,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5234440307014211,
+      "grad_norm": 2.0461785793304443,
+      "learning_rate": 0.00015,
+      "loss": 4.012592315673828,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5237480051675659,
+      "grad_norm": 1.9542787075042725,
+      "learning_rate": 0.00015,
+      "loss": 3.7990925312042236,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5240519796337108,
+      "grad_norm": 1.7799928188323975,
+      "learning_rate": 0.00015,
+      "loss": 4.215954780578613,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5243559540998556,
+      "grad_norm": 1.8881672620773315,
+      "learning_rate": 0.00015,
+      "loss": 4.061239242553711,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5246599285660004,
+      "grad_norm": 1.9614671468734741,
+      "learning_rate": 0.00015,
+      "loss": 4.082991123199463,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5249639030321454,
+      "grad_norm": 3.1234982013702393,
+      "learning_rate": 0.00015,
+      "loss": 3.815361499786377,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5252678774982902,
+      "grad_norm": 1.7297325134277344,
+      "learning_rate": 0.00015,
+      "loss": 4.319180488586426,
+      "step": 1728
+    },
+    {
+      "epoch": 0.525571851964435,
+      "grad_norm": 1.7399725914001465,
+      "learning_rate": 0.00015,
+      "loss": 4.231525897979736,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5258758264305798,
+      "grad_norm": 2.252669334411621,
+      "learning_rate": 0.00015,
+      "loss": 4.168580055236816,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5261798008967247,
+      "grad_norm": 2.3161239624023438,
+      "learning_rate": 0.00015,
+      "loss": 4.00864315032959,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5264837753628695,
+      "grad_norm": 2.1063270568847656,
+      "learning_rate": 0.00015,
+      "loss": 4.4139790534973145,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5267877498290143,
+      "grad_norm": 1.8122577667236328,
+      "learning_rate": 0.00015,
+      "loss": 4.011260032653809,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5270917242951592,
+      "grad_norm": 2.477172374725342,
+      "learning_rate": 0.00015,
+      "loss": 3.943798780441284,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5273956987613041,
+      "grad_norm": 1.8047622442245483,
+      "learning_rate": 0.00015,
+      "loss": 4.100638389587402,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5276996732274489,
+      "grad_norm": 2.237138032913208,
+      "learning_rate": 0.00015,
+      "loss": 4.172605037689209,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5280036476935938,
+      "grad_norm": 2.0994081497192383,
+      "learning_rate": 0.00015,
+      "loss": 4.112621784210205,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5283076221597386,
+      "grad_norm": 2.229337215423584,
+      "learning_rate": 0.00015,
+      "loss": 3.9645884037017822,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5286115966258834,
+      "grad_norm": 2.625911235809326,
+      "learning_rate": 0.00015,
+      "loss": 4.45831298828125,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5289155710920282,
+      "grad_norm": 2.2150564193725586,
+      "learning_rate": 0.00015,
+      "loss": 4.138443946838379,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5292195455581731,
+      "grad_norm": 2.010540723800659,
+      "learning_rate": 0.00015,
+      "loss": 3.874439001083374,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5295235200243179,
+      "grad_norm": 1.6998926401138306,
+      "learning_rate": 0.00015,
+      "loss": 3.927365303039551,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5298274944904628,
+      "grad_norm": 2.0175704956054688,
+      "learning_rate": 0.00015,
+      "loss": 4.049760341644287,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5301314689566077,
+      "grad_norm": 1.7769966125488281,
+      "learning_rate": 0.00015,
+      "loss": 3.8707919120788574,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5304354434227525,
+      "grad_norm": 2.334418535232544,
+      "learning_rate": 0.00015,
+      "loss": 4.249509811401367,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5307394178888973,
+      "grad_norm": 1.7730368375778198,
+      "learning_rate": 0.00015,
+      "loss": 3.982333183288574,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5310433923550422,
+      "grad_norm": 2.077176094055176,
+      "learning_rate": 0.00015,
+      "loss": 4.020873069763184,
+      "step": 1747
+    },
+    {
+      "epoch": 0.531347366821187,
+      "grad_norm": 1.8224660158157349,
+      "learning_rate": 0.00015,
+      "loss": 3.9774909019470215,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5316513412873318,
+      "grad_norm": 2.232837438583374,
+      "learning_rate": 0.00015,
+      "loss": 4.377743721008301,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5319553157534767,
+      "grad_norm": 1.935348391532898,
+      "learning_rate": 0.00015,
+      "loss": 3.919982433319092,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5322592902196216,
+      "grad_norm": 2.025601863861084,
+      "learning_rate": 0.00015,
+      "loss": 4.4491777420043945,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5325632646857664,
+      "grad_norm": 2.614570379257202,
+      "learning_rate": 0.00015,
+      "loss": 4.643579483032227,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5328672391519113,
+      "grad_norm": 1.9091076850891113,
+      "learning_rate": 0.00015,
+      "loss": 4.015471935272217,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5331712136180561,
+      "grad_norm": 1.9691740274429321,
+      "learning_rate": 0.00015,
+      "loss": 3.9966912269592285,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5334751880842009,
+      "grad_norm": 3.3222692012786865,
+      "learning_rate": 0.00015,
+      "loss": 3.8392183780670166,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5337791625503457,
+      "grad_norm": 2.679622173309326,
+      "learning_rate": 0.00015,
+      "loss": 4.596764087677002,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5340831370164906,
+      "grad_norm": 1.9135236740112305,
+      "learning_rate": 0.00015,
+      "loss": 3.831129550933838,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5343871114826355,
+      "grad_norm": 2.0269272327423096,
+      "learning_rate": 0.00015,
+      "loss": 4.055319309234619,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5346910859487803,
+      "grad_norm": 1.873222827911377,
+      "learning_rate": 0.00015,
+      "loss": 4.146714210510254,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5349950604149252,
+      "grad_norm": 1.8251131772994995,
+      "learning_rate": 0.00015,
+      "loss": 4.175183296203613,
+      "step": 1760
+    },
+    {
+      "epoch": 0.53529903488107,
+      "grad_norm": 2.0985589027404785,
+      "learning_rate": 0.00015,
+      "loss": 3.972984790802002,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5356030093472148,
+      "grad_norm": 2.1073696613311768,
+      "learning_rate": 0.00015,
+      "loss": 3.8077926635742188,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5359069838133597,
+      "grad_norm": 1.5445353984832764,
+      "learning_rate": 0.00015,
+      "loss": 4.051395416259766,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5362109582795045,
+      "grad_norm": 2.0442211627960205,
+      "learning_rate": 0.00015,
+      "loss": 4.110958576202393,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5365149327456493,
+      "grad_norm": 1.8371543884277344,
+      "learning_rate": 0.00015,
+      "loss": 3.804142475128174,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5368189072117943,
+      "grad_norm": 2.1567673683166504,
+      "learning_rate": 0.00015,
+      "loss": 4.000255584716797,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5371228816779391,
+      "grad_norm": 1.9297246932983398,
+      "learning_rate": 0.00015,
+      "loss": 4.191997528076172,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5374268561440839,
+      "grad_norm": 1.9444172382354736,
+      "learning_rate": 0.00015,
+      "loss": 4.214323043823242,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5377308306102287,
+      "grad_norm": 2.134854555130005,
+      "learning_rate": 0.00015,
+      "loss": 4.122425556182861,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5380348050763736,
+      "grad_norm": 1.9914805889129639,
+      "learning_rate": 0.00015,
+      "loss": 4.0743255615234375,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5383387795425184,
+      "grad_norm": 1.6992428302764893,
+      "learning_rate": 0.00015,
+      "loss": 4.357956886291504,
+      "step": 1771
+    },
+    {
+      "epoch": 0.5386427540086632,
+      "grad_norm": 1.7626954317092896,
+      "learning_rate": 0.00015,
+      "loss": 3.8260507583618164,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5389467284748081,
+      "grad_norm": 1.4458423852920532,
+      "learning_rate": 0.00015,
+      "loss": 4.113370895385742,
+      "step": 1773
+    },
+    {
+      "epoch": 0.539250702940953,
+      "grad_norm": 1.9367338418960571,
+      "learning_rate": 0.00015,
+      "loss": 4.050226211547852,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5395546774070978,
+      "grad_norm": 1.6504371166229248,
+      "learning_rate": 0.00015,
+      "loss": 4.272388935089111,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5398586518732427,
+      "grad_norm": 1.5321162939071655,
+      "learning_rate": 0.00015,
+      "loss": 3.878124475479126,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5401626263393875,
+      "grad_norm": 3.343132734298706,
+      "learning_rate": 0.00015,
+      "loss": 4.206187725067139,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5404666008055323,
+      "grad_norm": 1.881395697593689,
+      "learning_rate": 0.00015,
+      "loss": 3.967921733856201,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5407705752716772,
+      "grad_norm": 1.7244268655776978,
+      "learning_rate": 0.00015,
+      "loss": 4.005974292755127,
+      "step": 1779
+    },
+    {
+      "epoch": 0.541074549737822,
+      "grad_norm": 1.6167058944702148,
+      "learning_rate": 0.00015,
+      "loss": 4.0361175537109375,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5413785242039668,
+      "grad_norm": 8.240731239318848,
+      "learning_rate": 0.00015,
+      "loss": 4.453734397888184,
+      "step": 1781
+    },
+    {
+      "epoch": 0.5416824986701118,
+      "grad_norm": 2.7031474113464355,
+      "learning_rate": 0.00015,
+      "loss": 3.7253777980804443,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5419864731362566,
+      "grad_norm": 2.0473334789276123,
+      "learning_rate": 0.00015,
+      "loss": 4.113070964813232,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5422904476024014,
+      "grad_norm": 2.593289375305176,
+      "learning_rate": 0.00015,
+      "loss": 4.102260589599609,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5425944220685462,
+      "grad_norm": 1.8858399391174316,
+      "learning_rate": 0.00015,
+      "loss": 4.018864631652832,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5428983965346911,
+      "grad_norm": 1.7296861410140991,
+      "learning_rate": 0.00015,
+      "loss": 3.8227057456970215,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5432023710008359,
+      "grad_norm": 1.768086314201355,
+      "learning_rate": 0.00015,
+      "loss": 3.7416224479675293,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5435063454669807,
+      "grad_norm": 2.297962188720703,
+      "learning_rate": 0.00015,
+      "loss": 3.9193594455718994,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5438103199331256,
+      "grad_norm": 1.9373284578323364,
+      "learning_rate": 0.00015,
+      "loss": 4.0653767585754395,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5441142943992705,
+      "grad_norm": 2.4762890338897705,
+      "learning_rate": 0.00015,
+      "loss": 4.405949592590332,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5444182688654153,
+      "grad_norm": 2.060249090194702,
+      "learning_rate": 0.00015,
+      "loss": 4.367643356323242,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5447222433315602,
+      "grad_norm": 2.031403064727783,
+      "learning_rate": 0.00015,
+      "loss": 4.223331451416016,
+      "step": 1792
+    },
+    {
+      "epoch": 0.545026217797705,
+      "grad_norm": 2.153907060623169,
+      "learning_rate": 0.00015,
+      "loss": 4.170934200286865,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5453301922638498,
+      "grad_norm": 2.19777250289917,
+      "learning_rate": 0.00015,
+      "loss": 4.144471168518066,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5456341667299947,
+      "grad_norm": 2.1282241344451904,
+      "learning_rate": 0.00015,
+      "loss": 3.98885440826416,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5459381411961395,
+      "grad_norm": 2.4860377311706543,
+      "learning_rate": 0.00015,
+      "loss": 4.282989978790283,
+      "step": 1796
+    },
+    {
+      "epoch": 0.5462421156622844,
+      "grad_norm": 2.158522367477417,
+      "learning_rate": 0.00015,
+      "loss": 3.5447816848754883,
+      "step": 1797
+    },
+    {
+      "epoch": 0.5465460901284293,
+      "grad_norm": 1.7748351097106934,
+      "learning_rate": 0.00015,
+      "loss": 3.935945510864258,
+      "step": 1798
+    },
+    {
+      "epoch": 0.5468500645945741,
+      "grad_norm": 3.3034286499023438,
+      "learning_rate": 0.00015,
+      "loss": 4.149104118347168,
+      "step": 1799
+    },
+    {
+      "epoch": 0.5471540390607189,
+      "grad_norm": 1.9673985242843628,
+      "learning_rate": 0.00015,
+      "loss": 4.05586576461792,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5474580135268637,
+      "grad_norm": 1.5860862731933594,
+      "learning_rate": 0.00015,
+      "loss": 3.784270763397217,
+      "step": 1801
+    },
+    {
+      "epoch": 0.5477619879930086,
+      "grad_norm": 2.243295431137085,
+      "learning_rate": 0.00015,
+      "loss": 4.509024620056152,
+      "step": 1802
+    },
+    {
+      "epoch": 0.5480659624591534,
+      "grad_norm": 2.3135554790496826,
+      "learning_rate": 0.00015,
+      "loss": 3.680823802947998,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5483699369252982,
+      "grad_norm": 1.8965171575546265,
+      "learning_rate": 0.00015,
+      "loss": 4.23628568649292,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5486739113914432,
+      "grad_norm": 2.010997772216797,
+      "learning_rate": 0.00015,
+      "loss": 3.9184467792510986,
+      "step": 1805
+    },
+    {
+      "epoch": 0.548977885857588,
+      "grad_norm": 1.9178755283355713,
+      "learning_rate": 0.00015,
+      "loss": 3.945976495742798,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5492818603237328,
+      "grad_norm": 2.054088592529297,
+      "learning_rate": 0.00015,
+      "loss": 4.077604293823242,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5495858347898777,
+      "grad_norm": 3.5465476512908936,
+      "learning_rate": 0.00015,
+      "loss": 4.116716384887695,
+      "step": 1808
+    },
+    {
+      "epoch": 0.5498898092560225,
+      "grad_norm": 1.9903091192245483,
+      "learning_rate": 0.00015,
+      "loss": 4.23339319229126,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5501937837221673,
+      "grad_norm": 2.780412197113037,
+      "learning_rate": 0.00015,
+      "loss": 4.510908603668213,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5504977581883121,
+      "grad_norm": 1.903101921081543,
+      "learning_rate": 0.00015,
+      "loss": 4.352203845977783,
+      "step": 1811
+    },
+    {
+      "epoch": 0.550801732654457,
+      "grad_norm": 2.233325958251953,
+      "learning_rate": 0.00015,
+      "loss": 3.864790439605713,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5511057071206019,
+      "grad_norm": 1.8387113809585571,
+      "learning_rate": 0.00015,
+      "loss": 4.050426483154297,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5514096815867467,
+      "grad_norm": 1.966797947883606,
+      "learning_rate": 0.00015,
+      "loss": 4.2026214599609375,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5517136560528916,
+      "grad_norm": 2.033226728439331,
+      "learning_rate": 0.00015,
+      "loss": 3.881707191467285,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5520176305190364,
+      "grad_norm": 2.3971734046936035,
+      "learning_rate": 0.00015,
+      "loss": 3.737544298171997,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5523216049851812,
+      "grad_norm": 2.1885080337524414,
+      "learning_rate": 0.00015,
+      "loss": 4.204117774963379,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5526255794513261,
+      "grad_norm": 1.8974088430404663,
+      "learning_rate": 0.00015,
+      "loss": 4.078130722045898,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5529295539174709,
+      "grad_norm": 2.30381178855896,
+      "learning_rate": 0.00015,
+      "loss": 4.029746055603027,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5532335283836157,
+      "grad_norm": 2.620695114135742,
+      "learning_rate": 0.00015,
+      "loss": 3.8279528617858887,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5535375028497607,
+      "grad_norm": 2.090085506439209,
+      "learning_rate": 0.00015,
+      "loss": 4.220458984375,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5538414773159055,
+      "grad_norm": 2.0583314895629883,
+      "learning_rate": 0.00015,
+      "loss": 3.8411965370178223,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5541454517820503,
+      "grad_norm": 2.172788143157959,
+      "learning_rate": 0.00015,
+      "loss": 3.952096700668335,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5544494262481952,
+      "grad_norm": 1.97947096824646,
+      "learning_rate": 0.00015,
+      "loss": 3.9482784271240234,
+      "step": 1824
+    },
+    {
+      "epoch": 0.55475340071434,
+      "grad_norm": 2.0508415699005127,
+      "learning_rate": 0.00015,
+      "loss": 4.0625505447387695,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5550573751804848,
+      "grad_norm": 1.8764848709106445,
+      "learning_rate": 0.00015,
+      "loss": 3.885258913040161,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5553613496466296,
+      "grad_norm": 1.892353892326355,
+      "learning_rate": 0.00015,
+      "loss": 3.8676164150238037,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5556653241127745,
+      "grad_norm": 1.6236217021942139,
+      "learning_rate": 0.00015,
+      "loss": 3.9065160751342773,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5559692985789194,
+      "grad_norm": 1.6710917949676514,
+      "learning_rate": 0.00015,
+      "loss": 4.004802703857422,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5562732730450642,
+      "grad_norm": 2.3533897399902344,
+      "learning_rate": 0.00015,
+      "loss": 3.8975367546081543,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5565772475112091,
+      "grad_norm": 2.326199531555176,
+      "learning_rate": 0.00015,
+      "loss": 3.986532688140869,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5568812219773539,
+      "grad_norm": 1.9187307357788086,
+      "learning_rate": 0.00015,
+      "loss": 4.257449150085449,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5571851964434987,
+      "grad_norm": 1.7733986377716064,
+      "learning_rate": 0.00015,
+      "loss": 3.9229698181152344,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5574891709096436,
+      "grad_norm": 1.8732068538665771,
+      "learning_rate": 0.00015,
+      "loss": 3.9255495071411133,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5577931453757884,
+      "grad_norm": 2.031564950942993,
+      "learning_rate": 0.00015,
+      "loss": 4.222734451293945,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5580971198419333,
+      "grad_norm": 2.014289140701294,
+      "learning_rate": 0.00015,
+      "loss": 3.967153787612915,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5584010943080782,
+      "grad_norm": 2.2337281703948975,
+      "learning_rate": 0.00015,
+      "loss": 4.182089805603027,
+      "step": 1837
+    },
+    {
+      "epoch": 0.558705068774223,
+      "grad_norm": 1.9897671937942505,
+      "learning_rate": 0.00015,
+      "loss": 4.101202964782715,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5590090432403678,
+      "grad_norm": 1.955814242362976,
+      "learning_rate": 0.00015,
+      "loss": 4.060070991516113,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5593130177065126,
+      "grad_norm": 1.6479707956314087,
+      "learning_rate": 0.00015,
+      "loss": 3.750399112701416,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5596169921726575,
+      "grad_norm": 4.269290447235107,
+      "learning_rate": 0.00015,
+      "loss": 4.3132781982421875,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5599209666388023,
+      "grad_norm": 2.0882723331451416,
+      "learning_rate": 0.00015,
+      "loss": 4.197592258453369,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5602249411049471,
+      "grad_norm": 1.714022159576416,
+      "learning_rate": 0.00015,
+      "loss": 4.2926435470581055,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5605289155710921,
+      "grad_norm": 2.1894888877868652,
+      "learning_rate": 0.00015,
+      "loss": 3.872429847717285,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5608328900372369,
+      "grad_norm": 1.704744577407837,
+      "learning_rate": 0.00015,
+      "loss": 4.1356048583984375,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5611368645033817,
+      "grad_norm": 2.0460617542266846,
+      "learning_rate": 0.00015,
+      "loss": 4.132702827453613,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5614408389695266,
+      "grad_norm": 1.9006152153015137,
+      "learning_rate": 0.00015,
+      "loss": 3.8867645263671875,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5617448134356714,
+      "grad_norm": 2.292534589767456,
+      "learning_rate": 0.00015,
+      "loss": 4.148140907287598,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5620487879018162,
+      "grad_norm": 1.979906439781189,
+      "learning_rate": 0.00015,
+      "loss": 3.7265539169311523,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5623527623679611,
+      "grad_norm": 2.357837200164795,
+      "learning_rate": 0.00015,
+      "loss": 4.012168884277344,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5626567368341059,
+      "grad_norm": 2.1279752254486084,
+      "learning_rate": 0.00015,
+      "loss": 3.8114655017852783,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5629607113002508,
+      "grad_norm": 1.881948471069336,
+      "learning_rate": 0.00015,
+      "loss": 3.9761221408843994,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5632646857663957,
+      "grad_norm": 1.8558175563812256,
+      "learning_rate": 0.00015,
+      "loss": 3.9292421340942383,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5635686602325405,
+      "grad_norm": 1.9696072340011597,
+      "learning_rate": 0.00015,
+      "loss": 4.092259407043457,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5638726346986853,
+      "grad_norm": 2.0750677585601807,
+      "learning_rate": 0.00015,
+      "loss": 3.9417073726654053,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5641766091648301,
+      "grad_norm": 2.26316499710083,
+      "learning_rate": 0.00015,
+      "loss": 3.937424898147583,
+      "step": 1856
+    },
+    {
+      "epoch": 0.564480583630975,
+      "grad_norm": 2.245180368423462,
+      "learning_rate": 0.00015,
+      "loss": 3.808405876159668,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5647845580971198,
+      "grad_norm": 2.104825496673584,
+      "learning_rate": 0.00015,
+      "loss": 3.537917137145996,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5650885325632646,
+      "grad_norm": 1.9673336744308472,
+      "learning_rate": 0.00015,
+      "loss": 4.117812633514404,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5653925070294096,
+      "grad_norm": 1.8352693319320679,
+      "learning_rate": 0.00015,
+      "loss": 3.8555781841278076,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5656964814955544,
+      "grad_norm": 2.532613515853882,
+      "learning_rate": 0.00015,
+      "loss": 4.299034595489502,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5660004559616992,
+      "grad_norm": 1.8112268447875977,
+      "learning_rate": 0.00015,
+      "loss": 4.288586139678955,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5663044304278441,
+      "grad_norm": 1.9204347133636475,
+      "learning_rate": 0.00015,
+      "loss": 4.06419563293457,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5666084048939889,
+      "grad_norm": 2.3590915203094482,
+      "learning_rate": 0.00015,
+      "loss": 4.247583389282227,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5669123793601337,
+      "grad_norm": 2.0457310676574707,
+      "learning_rate": 0.00015,
+      "loss": 4.147165775299072,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5672163538262786,
+      "grad_norm": 1.993741512298584,
+      "learning_rate": 0.00015,
+      "loss": 4.109101295471191,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5675203282924234,
+      "grad_norm": 1.776987075805664,
+      "learning_rate": 0.00015,
+      "loss": 3.736893653869629,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5678243027585683,
+      "grad_norm": 1.8948038816452026,
+      "learning_rate": 0.00015,
+      "loss": 3.982174873352051,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5681282772247132,
+      "grad_norm": 2.158071517944336,
+      "learning_rate": 0.00015,
+      "loss": 3.728707790374756,
+      "step": 1869
+    },
+    {
+      "epoch": 0.568432251690858,
+      "grad_norm": 2.221409559249878,
+      "learning_rate": 0.00015,
+      "loss": 4.0619659423828125,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5687362261570028,
+      "grad_norm": 2.1257176399230957,
+      "learning_rate": 0.00015,
+      "loss": 3.9631333351135254,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5690402006231476,
+      "grad_norm": 1.8230996131896973,
+      "learning_rate": 0.00015,
+      "loss": 4.121955394744873,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5693441750892925,
+      "grad_norm": 2.027773141860962,
+      "learning_rate": 0.00015,
+      "loss": 3.9308042526245117,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5696481495554373,
+      "grad_norm": 1.813022255897522,
+      "learning_rate": 0.00015,
+      "loss": 4.024329662322998,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5699521240215822,
+      "grad_norm": 1.7805774211883545,
+      "learning_rate": 0.00015,
+      "loss": 4.19598388671875,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5702560984877271,
+      "grad_norm": 2.1109330654144287,
+      "learning_rate": 0.00015,
+      "loss": 4.053560256958008,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5705600729538719,
+      "grad_norm": 1.9273124933242798,
+      "learning_rate": 0.00015,
+      "loss": 4.097064971923828,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5708640474200167,
+      "grad_norm": 1.751803994178772,
+      "learning_rate": 0.00015,
+      "loss": 3.882071018218994,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5711680218861616,
+      "grad_norm": 2.386807918548584,
+      "learning_rate": 0.00015,
+      "loss": 3.7340729236602783,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5714719963523064,
+      "grad_norm": 1.687360167503357,
+      "learning_rate": 0.00015,
+      "loss": 3.8952767848968506,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5717759708184512,
+      "grad_norm": 2.2741405963897705,
+      "learning_rate": 0.00015,
+      "loss": 3.5224342346191406,
+      "step": 1881
+    },
+    {
+      "epoch": 0.572079945284596,
+      "grad_norm": 1.8575007915496826,
+      "learning_rate": 0.00015,
+      "loss": 4.0964508056640625,
+      "step": 1882
+    },
+    {
+      "epoch": 0.572383919750741,
+      "grad_norm": 2.0738608837127686,
+      "learning_rate": 0.00015,
+      "loss": 4.000520706176758,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5726878942168858,
+      "grad_norm": 1.90554678440094,
+      "learning_rate": 0.00015,
+      "loss": 3.9827537536621094,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5729918686830306,
+      "grad_norm": 1.8592333793640137,
+      "learning_rate": 0.00015,
+      "loss": 4.119943141937256,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5732958431491755,
+      "grad_norm": 1.8441842794418335,
+      "learning_rate": 0.00015,
+      "loss": 3.8373146057128906,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5735998176153203,
+      "grad_norm": 1.625861406326294,
+      "learning_rate": 0.00015,
+      "loss": 4.290899276733398,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5739037920814651,
+      "grad_norm": 1.7842477560043335,
+      "learning_rate": 0.00015,
+      "loss": 3.9913434982299805,
+      "step": 1888
+    },
+    {
+      "epoch": 0.57420776654761,
+      "grad_norm": 3.500227212905884,
+      "learning_rate": 0.00015,
+      "loss": 4.44783878326416,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5745117410137548,
+      "grad_norm": 1.7448029518127441,
+      "learning_rate": 0.00015,
+      "loss": 3.711030960083008,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5748157154798997,
+      "grad_norm": 1.6997615098953247,
+      "learning_rate": 0.00015,
+      "loss": 4.050522327423096,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5751196899460446,
+      "grad_norm": 2.1973581314086914,
+      "learning_rate": 0.00015,
+      "loss": 3.8546783924102783,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5754236644121894,
+      "grad_norm": 1.7385939359664917,
+      "learning_rate": 0.00015,
+      "loss": 4.00539493560791,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5757276388783342,
+      "grad_norm": 2.416806936264038,
+      "learning_rate": 0.00015,
+      "loss": 4.184887886047363,
+      "step": 1894
+    },
+    {
+      "epoch": 0.576031613344479,
+      "grad_norm": 1.926730751991272,
+      "learning_rate": 0.00015,
+      "loss": 3.774869441986084,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5763355878106239,
+      "grad_norm": 1.9684704542160034,
+      "learning_rate": 0.00015,
+      "loss": 4.124790191650391,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5766395622767687,
+      "grad_norm": 1.80181086063385,
+      "learning_rate": 0.00015,
+      "loss": 3.651839256286621,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5769435367429135,
+      "grad_norm": 2.039480209350586,
+      "learning_rate": 0.00015,
+      "loss": 3.799783706665039,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5772475112090585,
+      "grad_norm": 2.014221668243408,
+      "learning_rate": 0.00015,
+      "loss": 3.8968350887298584,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5775514856752033,
+      "grad_norm": 2.111248731613159,
+      "learning_rate": 0.00015,
+      "loss": 4.1519317626953125,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5778554601413481,
+      "grad_norm": 1.8858197927474976,
+      "learning_rate": 0.00015,
+      "loss": 3.903240203857422,
+      "step": 1901
+    },
+    {
+      "epoch": 0.578159434607493,
+      "grad_norm": 2.106196403503418,
+      "learning_rate": 0.00015,
+      "loss": 4.094615936279297,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5784634090736378,
+      "grad_norm": 1.7524354457855225,
+      "learning_rate": 0.00015,
+      "loss": 3.918078899383545,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5787673835397826,
+      "grad_norm": 2.0072553157806396,
+      "learning_rate": 0.00015,
+      "loss": 4.013356685638428,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5790713580059275,
+      "grad_norm": 2.1023335456848145,
+      "learning_rate": 0.00015,
+      "loss": 3.9322867393493652,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5793753324720723,
+      "grad_norm": 2.801133871078491,
+      "learning_rate": 0.00015,
+      "loss": 4.133281707763672,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5796793069382172,
+      "grad_norm": 2.297572612762451,
+      "learning_rate": 0.00015,
+      "loss": 3.9275858402252197,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5799832814043621,
+      "grad_norm": 1.9030765295028687,
+      "learning_rate": 0.00015,
+      "loss": 3.8344101905822754,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5802872558705069,
+      "grad_norm": 1.9185166358947754,
+      "learning_rate": 0.00015,
+      "loss": 4.531183242797852,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5805912303366517,
+      "grad_norm": 1.6808274984359741,
+      "learning_rate": 0.00015,
+      "loss": 4.144586086273193,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5808952048027965,
+      "grad_norm": 1.6830376386642456,
+      "learning_rate": 0.00015,
+      "loss": 3.9942264556884766,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5811991792689414,
+      "grad_norm": 1.6700451374053955,
+      "learning_rate": 0.00015,
+      "loss": 3.9169912338256836,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5815031537350862,
+      "grad_norm": 1.9618070125579834,
+      "learning_rate": 0.00015,
+      "loss": 3.903103828430176,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5818071282012311,
+      "grad_norm": 1.9430345296859741,
+      "learning_rate": 0.00015,
+      "loss": 3.6323394775390625,
+      "step": 1914
+    },
+    {
+      "epoch": 0.582111102667376,
+      "grad_norm": 2.1944591999053955,
+      "learning_rate": 0.00015,
+      "loss": 3.9079933166503906,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5824150771335208,
+      "grad_norm": 2.005342960357666,
+      "learning_rate": 0.00015,
+      "loss": 4.396800994873047,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5827190515996656,
+      "grad_norm": 1.8391380310058594,
+      "learning_rate": 0.00015,
+      "loss": 3.7923474311828613,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5830230260658105,
+      "grad_norm": 2.0687265396118164,
+      "learning_rate": 0.00015,
+      "loss": 4.032331466674805,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5833270005319553,
+      "grad_norm": 1.709262728691101,
+      "learning_rate": 0.00015,
+      "loss": 4.076603412628174,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5836309749981001,
+      "grad_norm": 2.067333459854126,
+      "learning_rate": 0.00015,
+      "loss": 4.255692481994629,
+      "step": 1920
+    },
+    {
+      "epoch": 0.583934949464245,
+      "grad_norm": 2.059882640838623,
+      "learning_rate": 0.00015,
+      "loss": 3.8534927368164062,
+      "step": 1921
+    },
+    {
+      "epoch": 0.5842389239303899,
+      "grad_norm": 1.9531704187393188,
+      "learning_rate": 0.00015,
+      "loss": 4.038595199584961,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5845428983965347,
+      "grad_norm": 1.6714838743209839,
+      "learning_rate": 0.00015,
+      "loss": 4.232665061950684,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5848468728626796,
+      "grad_norm": 2.3071298599243164,
+      "learning_rate": 0.00015,
+      "loss": 3.899740219116211,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5851508473288244,
+      "grad_norm": 1.9869287014007568,
+      "learning_rate": 0.00015,
+      "loss": 4.044865608215332,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5854548217949692,
+      "grad_norm": 2.1357157230377197,
+      "learning_rate": 0.00015,
+      "loss": 3.9323055744171143,
+      "step": 1926
+    },
+    {
+      "epoch": 0.585758796261114,
+      "grad_norm": 2.114234447479248,
+      "learning_rate": 0.00015,
+      "loss": 3.9222769737243652,
+      "step": 1927
+    },
+    {
+      "epoch": 0.5860627707272589,
+      "grad_norm": 2.4386401176452637,
+      "learning_rate": 0.00015,
+      "loss": 4.243875026702881,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5863667451934037,
+      "grad_norm": 2.190553665161133,
+      "learning_rate": 0.00015,
+      "loss": 4.253232955932617,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5866707196595486,
+      "grad_norm": 2.1068243980407715,
+      "learning_rate": 0.00015,
+      "loss": 4.1498918533325195,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5869746941256935,
+      "grad_norm": 1.933572769165039,
+      "learning_rate": 0.00015,
+      "loss": 3.9688339233398438,
+      "step": 1931
+    },
+    {
+      "epoch": 0.5872786685918383,
+      "grad_norm": 3.475432872772217,
+      "learning_rate": 0.00015,
+      "loss": 3.9919214248657227,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5875826430579831,
+      "grad_norm": 2.2731399536132812,
+      "learning_rate": 0.00015,
+      "loss": 3.812394618988037,
+      "step": 1933
+    },
+    {
+      "epoch": 0.587886617524128,
+      "grad_norm": 1.9664931297302246,
+      "learning_rate": 0.00015,
+      "loss": 3.8607592582702637,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5881905919902728,
+      "grad_norm": 2.096271276473999,
+      "learning_rate": 0.00015,
+      "loss": 3.7912561893463135,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5884945664564176,
+      "grad_norm": 1.9693447351455688,
+      "learning_rate": 0.00015,
+      "loss": 3.879939556121826,
+      "step": 1936
+    },
+    {
+      "epoch": 0.5887985409225625,
+      "grad_norm": 2.1325936317443848,
+      "learning_rate": 0.00015,
+      "loss": 3.8389031887054443,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5891025153887074,
+      "grad_norm": 1.9046918153762817,
+      "learning_rate": 0.00015,
+      "loss": 4.027096271514893,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5894064898548522,
+      "grad_norm": 1.8800222873687744,
+      "learning_rate": 0.00015,
+      "loss": 3.861910343170166,
+      "step": 1939
+    },
+    {
+      "epoch": 0.589710464320997,
+      "grad_norm": 1.7564054727554321,
+      "learning_rate": 0.00015,
+      "loss": 3.830658435821533,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5900144387871419,
+      "grad_norm": 2.8823599815368652,
+      "learning_rate": 0.00015,
+      "loss": 3.716590404510498,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5903184132532867,
+      "grad_norm": 2.0103771686553955,
+      "learning_rate": 0.00015,
+      "loss": 3.9385643005371094,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5906223877194315,
+      "grad_norm": 1.9599297046661377,
+      "learning_rate": 0.00015,
+      "loss": 3.952876567840576,
+      "step": 1943
+    },
+    {
+      "epoch": 0.5909263621855764,
+      "grad_norm": 1.7078571319580078,
+      "learning_rate": 0.00015,
+      "loss": 4.276788711547852,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5912303366517212,
+      "grad_norm": 1.9141947031021118,
+      "learning_rate": 0.00015,
+      "loss": 3.904168128967285,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5915343111178661,
+      "grad_norm": 2.5790789127349854,
+      "learning_rate": 0.00015,
+      "loss": 4.008854389190674,
+      "step": 1946
+    },
+    {
+      "epoch": 0.591838285584011,
+      "grad_norm": 2.400451421737671,
+      "learning_rate": 0.00015,
+      "loss": 3.746690273284912,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5921422600501558,
+      "grad_norm": 2.0254886150360107,
+      "learning_rate": 0.00015,
+      "loss": 3.9503588676452637,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5924462345163006,
+      "grad_norm": 1.8135665655136108,
+      "learning_rate": 0.00015,
+      "loss": 3.9519853591918945,
+      "step": 1949
+    },
+    {
+      "epoch": 0.5927502089824455,
+      "grad_norm": 2.13578462600708,
+      "learning_rate": 0.00015,
+      "loss": 3.8456945419311523,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5930541834485903,
+      "grad_norm": 1.6836427450180054,
+      "learning_rate": 0.00015,
+      "loss": 3.936838150024414,
+      "step": 1951
+    },
+    {
+      "epoch": 0.5933581579147351,
+      "grad_norm": 2.0635225772857666,
+      "learning_rate": 0.00015,
+      "loss": 4.189089298248291,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5936621323808801,
+      "grad_norm": 2.190361261367798,
+      "learning_rate": 0.00015,
+      "loss": 3.792356491088867,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5939661068470249,
+      "grad_norm": 1.762553334236145,
+      "learning_rate": 0.00015,
+      "loss": 4.141273498535156,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5942700813131697,
+      "grad_norm": 2.4575345516204834,
+      "learning_rate": 0.00015,
+      "loss": 4.410109519958496,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5945740557793145,
+      "grad_norm": 1.991228461265564,
+      "learning_rate": 0.00015,
+      "loss": 3.7756195068359375,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5948780302454594,
+      "grad_norm": 2.457785129547119,
+      "learning_rate": 0.00015,
+      "loss": 3.9305405616760254,
+      "step": 1957
+    },
+    {
+      "epoch": 0.5951820047116042,
+      "grad_norm": 2.1926310062408447,
+      "learning_rate": 0.00015,
+      "loss": 3.978116273880005,
+      "step": 1958
+    },
+    {
+      "epoch": 0.595485979177749,
+      "grad_norm": 1.9880603551864624,
+      "learning_rate": 0.00015,
+      "loss": 3.9303958415985107,
+      "step": 1959
+    },
+    {
+      "epoch": 0.5957899536438939,
+      "grad_norm": 2.050541400909424,
+      "learning_rate": 0.00015,
+      "loss": 3.9194552898406982,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5960939281100388,
+      "grad_norm": 2.1796367168426514,
+      "learning_rate": 0.00015,
+      "loss": 4.279801368713379,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5963979025761836,
+      "grad_norm": 2.190380573272705,
+      "learning_rate": 0.00015,
+      "loss": 4.106325149536133,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5967018770423285,
+      "grad_norm": 1.8914604187011719,
+      "learning_rate": 0.00015,
+      "loss": 3.823380470275879,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5970058515084733,
+      "grad_norm": 2.2076947689056396,
+      "learning_rate": 0.00015,
+      "loss": 3.7468271255493164,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5973098259746181,
+      "grad_norm": 2.083728313446045,
+      "learning_rate": 0.00015,
+      "loss": 4.082466125488281,
+      "step": 1965
+    },
+    {
+      "epoch": 0.597613800440763,
+      "grad_norm": 1.7240725755691528,
+      "learning_rate": 0.00015,
+      "loss": 3.7918758392333984,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5979177749069078,
+      "grad_norm": 1.9857996702194214,
+      "learning_rate": 0.00015,
+      "loss": 3.7307512760162354,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5982217493730526,
+      "grad_norm": 2.2352609634399414,
+      "learning_rate": 0.00015,
+      "loss": 3.864438056945801,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5985257238391976,
+      "grad_norm": 1.9411993026733398,
+      "learning_rate": 0.00015,
+      "loss": 4.020608901977539,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5988296983053424,
+      "grad_norm": 2.2522366046905518,
+      "learning_rate": 0.00015,
+      "loss": 4.145511627197266,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5991336727714872,
+      "grad_norm": 18.407054901123047,
+      "learning_rate": 0.00015,
+      "loss": 3.892716407775879,
+      "step": 1971
+    },
+    {
+      "epoch": 0.599437647237632,
+      "grad_norm": 1.7937145233154297,
+      "learning_rate": 0.00015,
+      "loss": 3.8897716999053955,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5997416217037769,
+      "grad_norm": 2.1579720973968506,
+      "learning_rate": 0.00015,
+      "loss": 4.1412858963012695,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6000455961699217,
+      "grad_norm": 2.0194270610809326,
+      "learning_rate": 0.00015,
+      "loss": 4.448011875152588,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6003495706360665,
+      "grad_norm": 2.4343035221099854,
+      "learning_rate": 0.00015,
+      "loss": 4.16526985168457,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6006535451022114,
+      "grad_norm": 2.314803123474121,
+      "learning_rate": 0.00015,
+      "loss": 3.8915932178497314,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6009575195683563,
+      "grad_norm": 1.9705531597137451,
+      "learning_rate": 0.00015,
+      "loss": 3.7638792991638184,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6012614940345011,
+      "grad_norm": 2.0214040279388428,
+      "learning_rate": 0.00015,
+      "loss": 4.073162078857422,
+      "step": 1978
+    },
+    {
+      "epoch": 0.601565468500646,
+      "grad_norm": 1.7325557470321655,
+      "learning_rate": 0.00015,
+      "loss": 3.8510499000549316,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6018694429667908,
+      "grad_norm": 1.8389602899551392,
+      "learning_rate": 0.00015,
+      "loss": 3.764779567718506,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6021734174329356,
+      "grad_norm": 1.8025801181793213,
+      "learning_rate": 0.00015,
+      "loss": 3.559173822402954,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6024773918990804,
+      "grad_norm": 1.8269233703613281,
+      "learning_rate": 0.00015,
+      "loss": 3.607572078704834,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6027813663652253,
+      "grad_norm": 1.9028425216674805,
+      "learning_rate": 0.00015,
+      "loss": 3.8880534172058105,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6030853408313702,
+      "grad_norm": 2.5928027629852295,
+      "learning_rate": 0.00015,
+      "loss": 3.804825782775879,
+      "step": 1984
+    },
+    {
+      "epoch": 0.603389315297515,
+      "grad_norm": 2.2426111698150635,
+      "learning_rate": 0.00015,
+      "loss": 4.294267654418945,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6036932897636599,
+      "grad_norm": 2.292942523956299,
+      "learning_rate": 0.00015,
+      "loss": 4.024377346038818,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6039972642298047,
+      "grad_norm": 2.302459955215454,
+      "learning_rate": 0.00015,
+      "loss": 4.094934463500977,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6043012386959495,
+      "grad_norm": 1.939491629600525,
+      "learning_rate": 0.00015,
+      "loss": 3.765880823135376,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6046052131620944,
+      "grad_norm": 2.5349888801574707,
+      "learning_rate": 0.00015,
+      "loss": 4.009184837341309,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6049091876282392,
+      "grad_norm": 2.0671913623809814,
+      "learning_rate": 0.00015,
+      "loss": 3.771256446838379,
+      "step": 1990
+    },
+    {
+      "epoch": 0.605213162094384,
+      "grad_norm": 2.4364826679229736,
+      "learning_rate": 0.00015,
+      "loss": 3.6627604961395264,
+      "step": 1991
+    },
+    {
+      "epoch": 0.605517136560529,
+      "grad_norm": 2.637615203857422,
+      "learning_rate": 0.00015,
+      "loss": 3.976627826690674,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6058211110266738,
+      "grad_norm": 1.9688924551010132,
+      "learning_rate": 0.00015,
+      "loss": 3.811709403991699,
+      "step": 1993
+    },
+    {
+      "epoch": 0.6061250854928186,
+      "grad_norm": 2.9559080600738525,
+      "learning_rate": 0.00015,
+      "loss": 3.5838308334350586,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6064290599589635,
+      "grad_norm": 1.887712836265564,
+      "learning_rate": 0.00015,
+      "loss": 3.879333019256592,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6067330344251083,
+      "grad_norm": 2.043086290359497,
+      "learning_rate": 0.00015,
+      "loss": 4.0009660720825195,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6070370088912531,
+      "grad_norm": 2.15146803855896,
+      "learning_rate": 0.00015,
+      "loss": 4.149009704589844,
+      "step": 1997
+    },
+    {
+      "epoch": 0.607340983357398,
+      "grad_norm": 2.1712963581085205,
+      "learning_rate": 0.00015,
+      "loss": 3.9999687671661377,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6076449578235428,
+      "grad_norm": 2.6381912231445312,
+      "learning_rate": 0.00015,
+      "loss": 3.848398208618164,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6079489322896877,
+      "grad_norm": 2.124262571334839,
+      "learning_rate": 0.00015,
+      "loss": 3.853292942047119,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6082529067558325,
+      "grad_norm": 1.8373878002166748,
+      "learning_rate": 0.00015,
+      "loss": 3.7545626163482666,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6085568812219774,
+      "grad_norm": 2.0599775314331055,
+      "learning_rate": 0.00015,
+      "loss": 4.322625637054443,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6088608556881222,
+      "grad_norm": 1.7490202188491821,
+      "learning_rate": 0.00015,
+      "loss": 3.9014856815338135,
+      "step": 2003
+    },
+    {
+      "epoch": 0.609164830154267,
+      "grad_norm": 2.1702680587768555,
+      "learning_rate": 0.00015,
+      "loss": 3.8320136070251465,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6094688046204119,
+      "grad_norm": 2.252533197402954,
+      "learning_rate": 0.00015,
+      "loss": 3.8732757568359375,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6097727790865567,
+      "grad_norm": 1.9656338691711426,
+      "learning_rate": 0.00015,
+      "loss": 3.9795186519622803,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6100767535527015,
+      "grad_norm": 2.2128653526306152,
+      "learning_rate": 0.00015,
+      "loss": 3.7235095500946045,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6103807280188465,
+      "grad_norm": 2.0747570991516113,
+      "learning_rate": 0.00015,
+      "loss": 3.9155163764953613,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6106847024849913,
+      "grad_norm": 1.8632644414901733,
+      "learning_rate": 0.00015,
+      "loss": 4.144404411315918,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6109886769511361,
+      "grad_norm": 1.9575724601745605,
+      "learning_rate": 0.00015,
+      "loss": 3.8008365631103516,
+      "step": 2010
+    },
+    {
+      "epoch": 0.611292651417281,
+      "grad_norm": 1.903622031211853,
+      "learning_rate": 0.00015,
+      "loss": 3.618894338607788,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6115966258834258,
+      "grad_norm": 2.182652473449707,
+      "learning_rate": 0.00015,
+      "loss": 3.7469663619995117,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6119006003495706,
+      "grad_norm": 1.8537768125534058,
+      "learning_rate": 0.00015,
+      "loss": 3.9289731979370117,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6122045748157154,
+      "grad_norm": 2.502333164215088,
+      "learning_rate": 0.00015,
+      "loss": 3.771933078765869,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6125085492818603,
+      "grad_norm": 1.6783372163772583,
+      "learning_rate": 0.00015,
+      "loss": 4.2244744300842285,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6128125237480052,
+      "grad_norm": 1.7165749073028564,
+      "learning_rate": 0.00015,
+      "loss": 4.071516513824463,
+      "step": 2016
+    },
+    {
+      "epoch": 0.61311649821415,
+      "grad_norm": 2.626720905303955,
+      "learning_rate": 0.00015,
+      "loss": 3.8607029914855957,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6134204726802949,
+      "grad_norm": 1.9258846044540405,
+      "learning_rate": 0.00015,
+      "loss": 4.147579669952393,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6137244471464397,
+      "grad_norm": 1.7497175931930542,
+      "learning_rate": 0.00015,
+      "loss": 4.064929485321045,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6140284216125845,
+      "grad_norm": 1.894215703010559,
+      "learning_rate": 0.00015,
+      "loss": 3.882809638977051,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6143323960787294,
+      "grad_norm": 8.122392654418945,
+      "learning_rate": 0.00015,
+      "loss": 3.7307300567626953,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6146363705448742,
+      "grad_norm": 2.3993184566497803,
+      "learning_rate": 0.00015,
+      "loss": 3.9094648361206055,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6149403450110191,
+      "grad_norm": 2.127119779586792,
+      "learning_rate": 0.00015,
+      "loss": 3.725557804107666,
+      "step": 2023
+    },
+    {
+      "epoch": 0.615244319477164,
+      "grad_norm": 1.6872144937515259,
+      "learning_rate": 0.00015,
+      "loss": 3.833655595779419,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6155482939433088,
+      "grad_norm": 1.7414679527282715,
+      "learning_rate": 0.00015,
+      "loss": 3.9519214630126953,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6158522684094536,
+      "grad_norm": 1.7396965026855469,
+      "learning_rate": 0.00015,
+      "loss": 4.012292861938477,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6161562428755984,
+      "grad_norm": 1.6447455883026123,
+      "learning_rate": 0.00015,
+      "loss": 3.877077579498291,
+      "step": 2027
+    },
+    {
+      "epoch": 0.6164602173417433,
+      "grad_norm": 2.1188552379608154,
+      "learning_rate": 0.00015,
+      "loss": 3.764613389968872,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6167641918078881,
+      "grad_norm": 2.127617597579956,
+      "learning_rate": 0.00015,
+      "loss": 4.192755699157715,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6170681662740329,
+      "grad_norm": 1.6660685539245605,
+      "learning_rate": 0.00015,
+      "loss": 3.8628809452056885,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6173721407401779,
+      "grad_norm": 2.4943296909332275,
+      "learning_rate": 0.00015,
+      "loss": 3.949005126953125,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6176761152063227,
+      "grad_norm": 1.8171589374542236,
+      "learning_rate": 0.00015,
+      "loss": 4.12439489364624,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6179800896724675,
+      "grad_norm": 1.783606767654419,
+      "learning_rate": 0.00015,
+      "loss": 4.135004043579102,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6182840641386124,
+      "grad_norm": 1.715234398841858,
+      "learning_rate": 0.00015,
+      "loss": 4.330992698669434,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6185880386047572,
+      "grad_norm": 1.9551708698272705,
+      "learning_rate": 0.00015,
+      "loss": 4.060796737670898,
+      "step": 2035
+    },
+    {
+      "epoch": 0.618892013070902,
+      "grad_norm": 2.153911590576172,
+      "learning_rate": 0.00015,
+      "loss": 3.9279565811157227,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6191959875370469,
+      "grad_norm": 2.014220714569092,
+      "learning_rate": 0.00015,
+      "loss": 4.110243797302246,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6194999620031917,
+      "grad_norm": 1.8282814025878906,
+      "learning_rate": 0.00015,
+      "loss": 3.976031541824341,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6198039364693366,
+      "grad_norm": 1.7598329782485962,
+      "learning_rate": 0.00015,
+      "loss": 3.6389517784118652,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6201079109354815,
+      "grad_norm": 1.784008264541626,
+      "learning_rate": 0.00015,
+      "loss": 4.00566291809082,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6204118854016263,
+      "grad_norm": 1.8213446140289307,
+      "learning_rate": 0.00015,
+      "loss": 3.869266986846924,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6207158598677711,
+      "grad_norm": 2.255573034286499,
+      "learning_rate": 0.00015,
+      "loss": 3.9049744606018066,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6210198343339159,
+      "grad_norm": 1.9965128898620605,
+      "learning_rate": 0.00015,
+      "loss": 3.7763466835021973,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6213238088000608,
+      "grad_norm": 2.1900601387023926,
+      "learning_rate": 0.00015,
+      "loss": 4.098472595214844,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6216277832662056,
+      "grad_norm": 1.955759048461914,
+      "learning_rate": 0.00015,
+      "loss": 4.615511894226074,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6219317577323504,
+      "grad_norm": 1.6639600992202759,
+      "learning_rate": 0.00015,
+      "loss": 3.9803786277770996,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6222357321984954,
+      "grad_norm": 2.3109898567199707,
+      "learning_rate": 0.00015,
+      "loss": 4.2724199295043945,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6225397066646402,
+      "grad_norm": 1.7768925428390503,
+      "learning_rate": 0.00015,
+      "loss": 4.036463737487793,
+      "step": 2048
+    },
+    {
+      "epoch": 0.622843681130785,
+      "grad_norm": 2.0974133014678955,
+      "learning_rate": 0.00015,
+      "loss": 4.072963714599609,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6231476555969299,
+      "grad_norm": 2.371058702468872,
+      "learning_rate": 0.00015,
+      "loss": 3.6987733840942383,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6234516300630747,
+      "grad_norm": 1.9242416620254517,
+      "learning_rate": 0.00015,
+      "loss": 3.9163029193878174,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6237556045292195,
+      "grad_norm": 2.15415620803833,
+      "learning_rate": 0.00015,
+      "loss": 3.7995176315307617,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6240595789953644,
+      "grad_norm": 2.2549870014190674,
+      "learning_rate": 0.00015,
+      "loss": 4.00458288192749,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6243635534615092,
+      "grad_norm": 2.8551206588745117,
+      "learning_rate": 0.00015,
+      "loss": 3.8119990825653076,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6246675279276541,
+      "grad_norm": 1.988860845565796,
+      "learning_rate": 0.00015,
+      "loss": 3.962386131286621,
+      "step": 2055
+    },
+    {
+      "epoch": 0.624971502393799,
+      "grad_norm": 1.7669918537139893,
+      "learning_rate": 0.00015,
+      "loss": 3.937694549560547,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6252754768599438,
+      "grad_norm": 2.0906083583831787,
+      "learning_rate": 0.00015,
+      "loss": 3.9988977909088135,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6255794513260886,
+      "grad_norm": 2.359090566635132,
+      "learning_rate": 0.00015,
+      "loss": 4.033588886260986,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6258834257922334,
+      "grad_norm": 2.194016218185425,
+      "learning_rate": 0.00015,
+      "loss": 3.913755178451538,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6261874002583783,
+      "grad_norm": 2.217017889022827,
+      "learning_rate": 0.00015,
+      "loss": 4.181171894073486,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6264913747245231,
+      "grad_norm": 1.7576452493667603,
+      "learning_rate": 0.00015,
+      "loss": 3.900646686553955,
+      "step": 2061
+    },
+    {
+      "epoch": 0.626795349190668,
+      "grad_norm": 2.0781922340393066,
+      "learning_rate": 0.00015,
+      "loss": 3.90639328956604,
+      "step": 2062
+    },
+    {
+      "epoch": 0.6270993236568129,
+      "grad_norm": 1.957253098487854,
+      "learning_rate": 0.00015,
+      "loss": 4.079903602600098,
+      "step": 2063
+    },
+    {
+      "epoch": 0.6274032981229577,
+      "grad_norm": 2.192580461502075,
+      "learning_rate": 0.00015,
+      "loss": 3.9237465858459473,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6277072725891025,
+      "grad_norm": 1.8959115743637085,
+      "learning_rate": 0.00015,
+      "loss": 4.270694732666016,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6280112470552474,
+      "grad_norm": 2.017061948776245,
+      "learning_rate": 0.00015,
+      "loss": 3.81596040725708,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6283152215213922,
+      "grad_norm": 2.068781852722168,
+      "learning_rate": 0.00015,
+      "loss": 3.9142072200775146,
+      "step": 2067
+    },
+    {
+      "epoch": 0.628619195987537,
+      "grad_norm": 2.028555393218994,
+      "learning_rate": 0.00015,
+      "loss": 3.7752556800842285,
+      "step": 2068
+    },
+    {
+      "epoch": 0.6289231704536818,
+      "grad_norm": 2.767035484313965,
+      "learning_rate": 0.00015,
+      "loss": 4.652556896209717,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6292271449198268,
+      "grad_norm": 2.8049020767211914,
+      "learning_rate": 0.00015,
+      "loss": 3.8813798427581787,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6295311193859716,
+      "grad_norm": 2.379903554916382,
+      "learning_rate": 0.00015,
+      "loss": 3.8963332176208496,
+      "step": 2071
+    },
+    {
+      "epoch": 0.6298350938521164,
+      "grad_norm": 2.8132965564727783,
+      "learning_rate": 0.00015,
+      "loss": 4.048893928527832,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6301390683182613,
+      "grad_norm": 2.153480052947998,
+      "learning_rate": 0.00015,
+      "loss": 3.595803737640381,
+      "step": 2073
+    },
+    {
+      "epoch": 0.6304430427844061,
+      "grad_norm": 2.08880615234375,
+      "learning_rate": 0.00015,
+      "loss": 3.470710277557373,
+      "step": 2074
+    },
+    {
+      "epoch": 0.6307470172505509,
+      "grad_norm": 1.5208089351654053,
+      "learning_rate": 0.00015,
+      "loss": 3.8215363025665283,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6310509917166958,
+      "grad_norm": 2.83600115776062,
+      "learning_rate": 0.00015,
+      "loss": 3.9152636528015137,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6313549661828406,
+      "grad_norm": 1.691405177116394,
+      "learning_rate": 0.00015,
+      "loss": 3.8861076831817627,
+      "step": 2077
+    },
+    {
+      "epoch": 0.6316589406489855,
+      "grad_norm": 1.838105320930481,
+      "learning_rate": 0.00015,
+      "loss": 4.056221961975098,
+      "step": 2078
+    },
+    {
+      "epoch": 0.6319629151151304,
+      "grad_norm": 2.623918294906616,
+      "learning_rate": 0.00015,
+      "loss": 3.7138125896453857,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6322668895812752,
+      "grad_norm": 1.906796932220459,
+      "learning_rate": 0.00015,
+      "loss": 4.045827865600586,
+      "step": 2080
+    },
+    {
+      "epoch": 0.63257086404742,
+      "grad_norm": 2.2463150024414062,
+      "learning_rate": 0.00015,
+      "loss": 3.8114171028137207,
+      "step": 2081
+    },
+    {
+      "epoch": 0.6328748385135649,
+      "grad_norm": 2.1478726863861084,
+      "learning_rate": 0.00015,
+      "loss": 3.9581527709960938,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6331788129797097,
+      "grad_norm": 2.397303819656372,
+      "learning_rate": 0.00015,
+      "loss": 3.5960147380828857,
+      "step": 2083
+    },
+    {
+      "epoch": 0.6334827874458545,
+      "grad_norm": 1.9836628437042236,
+      "learning_rate": 0.00015,
+      "loss": 3.6900033950805664,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6337867619119993,
+      "grad_norm": 2.4586944580078125,
+      "learning_rate": 0.00015,
+      "loss": 3.6679630279541016,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6340907363781443,
+      "grad_norm": 1.9627068042755127,
+      "learning_rate": 0.00015,
+      "loss": 3.914689540863037,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6343947108442891,
+      "grad_norm": 1.8647435903549194,
+      "learning_rate": 0.00015,
+      "loss": 3.5315916538238525,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6346986853104339,
+      "grad_norm": 2.310479164123535,
+      "learning_rate": 0.00015,
+      "loss": 3.8930182456970215,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6350026597765788,
+      "grad_norm": 1.7579567432403564,
+      "learning_rate": 0.00015,
+      "loss": 4.000404357910156,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6353066342427236,
+      "grad_norm": 1.6400961875915527,
+      "learning_rate": 0.00015,
+      "loss": 3.9366683959960938,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6356106087088684,
+      "grad_norm": 2.0321216583251953,
+      "learning_rate": 0.00015,
+      "loss": 4.1233110427856445,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6359145831750133,
+      "grad_norm": 3.572274684906006,
+      "learning_rate": 0.00015,
+      "loss": 4.008485794067383,
+      "step": 2092
+    },
+    {
+      "epoch": 0.6362185576411581,
+      "grad_norm": 2.2659919261932373,
+      "learning_rate": 0.00015,
+      "loss": 4.003087043762207,
+      "step": 2093
+    },
+    {
+      "epoch": 0.636522532107303,
+      "grad_norm": 2.097054958343506,
+      "learning_rate": 0.00015,
+      "loss": 4.043601036071777,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6368265065734479,
+      "grad_norm": 1.7065457105636597,
+      "learning_rate": 0.00015,
+      "loss": 3.806802749633789,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6371304810395927,
+      "grad_norm": 2.0900075435638428,
+      "learning_rate": 0.00015,
+      "loss": 3.8869409561157227,
+      "step": 2096
+    },
+    {
+      "epoch": 0.6374344555057375,
+      "grad_norm": 1.8590515851974487,
+      "learning_rate": 0.00015,
+      "loss": 3.9241690635681152,
+      "step": 2097
+    },
+    {
+      "epoch": 0.6377384299718823,
+      "grad_norm": 3.0407698154449463,
+      "learning_rate": 0.00015,
+      "loss": 4.302334308624268,
+      "step": 2098
+    },
+    {
+      "epoch": 0.6380424044380272,
+      "grad_norm": 2.0897152423858643,
+      "learning_rate": 0.00015,
+      "loss": 3.9782042503356934,
+      "step": 2099
+    },
+    {
+      "epoch": 0.638346378904172,
+      "grad_norm": 1.6587401628494263,
+      "learning_rate": 0.00015,
+      "loss": 3.964503765106201,
+      "step": 2100
+    },
+    {
+      "epoch": 0.638650353370317,
+      "grad_norm": 1.8034837245941162,
+      "learning_rate": 0.00015,
+      "loss": 4.1746320724487305,
+      "step": 2101
+    },
+    {
+      "epoch": 0.6389543278364618,
+      "grad_norm": 1.945542812347412,
+      "learning_rate": 0.00015,
+      "loss": 3.9585928916931152,
+      "step": 2102
+    },
+    {
+      "epoch": 0.6392583023026066,
+      "grad_norm": 1.9585648775100708,
+      "learning_rate": 0.00015,
+      "loss": 3.9822497367858887,
+      "step": 2103
+    },
+    {
+      "epoch": 0.6395622767687514,
+      "grad_norm": 2.134007215499878,
+      "learning_rate": 0.00015,
+      "loss": 3.9795336723327637,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6398662512348963,
+      "grad_norm": 2.005767583847046,
+      "learning_rate": 0.00015,
+      "loss": 3.724485397338867,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6401702257010411,
+      "grad_norm": 1.9242228269577026,
+      "learning_rate": 0.00015,
+      "loss": 3.991020441055298,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6404742001671859,
+      "grad_norm": 1.8499397039413452,
+      "learning_rate": 0.00015,
+      "loss": 3.8599486351013184,
+      "step": 2107
+    },
+    {
+      "epoch": 0.6407781746333308,
+      "grad_norm": 1.7520183324813843,
+      "learning_rate": 0.00015,
+      "loss": 3.743957996368408,
+      "step": 2108
+    },
+    {
+      "epoch": 0.6410821490994757,
+      "grad_norm": 2.206728935241699,
+      "learning_rate": 0.00015,
+      "loss": 3.7004270553588867,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6413861235656205,
+      "grad_norm": 1.8216842412948608,
+      "learning_rate": 0.00015,
+      "loss": 3.889678478240967,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6416900980317654,
+      "grad_norm": 2.59770131111145,
+      "learning_rate": 0.00015,
+      "loss": 4.0203447341918945,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6419940724979102,
+      "grad_norm": 1.8030059337615967,
+      "learning_rate": 0.00015,
+      "loss": 3.7178866863250732,
+      "step": 2112
+    },
+    {
+      "epoch": 0.642298046964055,
+      "grad_norm": 2.3276422023773193,
+      "learning_rate": 0.00015,
+      "loss": 4.131110191345215,
+      "step": 2113
+    },
+    {
+      "epoch": 0.6426020214301998,
+      "grad_norm": 2.2343904972076416,
+      "learning_rate": 0.00015,
+      "loss": 3.6767258644104004,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6429059958963447,
+      "grad_norm": 2.230315685272217,
+      "learning_rate": 0.00015,
+      "loss": 3.8168461322784424,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6432099703624895,
+      "grad_norm": 1.9215577840805054,
+      "learning_rate": 0.00015,
+      "loss": 3.829709529876709,
+      "step": 2116
+    },
+    {
+      "epoch": 0.6435139448286344,
+      "grad_norm": 7.045118808746338,
+      "learning_rate": 0.00015,
+      "loss": 4.4218363761901855,
+      "step": 2117
+    },
+    {
+      "epoch": 0.6438179192947793,
+      "grad_norm": 1.776617169380188,
+      "learning_rate": 0.00015,
+      "loss": 3.7728312015533447,
+      "step": 2118
+    },
+    {
+      "epoch": 0.6441218937609241,
+      "grad_norm": 2.036564350128174,
+      "learning_rate": 0.00015,
+      "loss": 3.6658308506011963,
+      "step": 2119
+    },
+    {
+      "epoch": 0.6444258682270689,
+      "grad_norm": 2.106139898300171,
+      "learning_rate": 0.00015,
+      "loss": 3.591310739517212,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6447298426932138,
+      "grad_norm": 1.8513054847717285,
+      "learning_rate": 0.00015,
+      "loss": 3.674602746963501,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6450338171593586,
+      "grad_norm": 2.735053539276123,
+      "learning_rate": 0.00015,
+      "loss": 3.709904432296753,
+      "step": 2122
+    },
+    {
+      "epoch": 0.6453377916255034,
+      "grad_norm": 2.117337465286255,
+      "learning_rate": 0.00015,
+      "loss": 3.883706569671631,
+      "step": 2123
+    },
+    {
+      "epoch": 0.6456417660916483,
+      "grad_norm": 2.275052785873413,
+      "learning_rate": 0.00015,
+      "loss": 3.8572428226470947,
+      "step": 2124
+    },
+    {
+      "epoch": 0.6459457405577932,
+      "grad_norm": 1.9843311309814453,
+      "learning_rate": 0.00015,
+      "loss": 3.8900837898254395,
+      "step": 2125
+    },
+    {
+      "epoch": 0.646249715023938,
+      "grad_norm": 1.951188564300537,
+      "learning_rate": 0.00015,
+      "loss": 3.919292688369751,
+      "step": 2126
+    },
+    {
+      "epoch": 0.6465536894900829,
+      "grad_norm": 1.916258454322815,
+      "learning_rate": 0.00015,
+      "loss": 3.967928886413574,
+      "step": 2127
+    },
+    {
+      "epoch": 0.6468576639562277,
+      "grad_norm": 1.8903229236602783,
+      "learning_rate": 0.00015,
+      "loss": 3.8180127143859863,
+      "step": 2128
+    },
+    {
+      "epoch": 0.6471616384223725,
+      "grad_norm": 2.0518476963043213,
+      "learning_rate": 0.00015,
+      "loss": 3.737178325653076,
+      "step": 2129
+    },
+    {
+      "epoch": 0.6474656128885173,
+      "grad_norm": 2.006810188293457,
+      "learning_rate": 0.00015,
+      "loss": 4.018744468688965,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6477695873546622,
+      "grad_norm": 2.0046606063842773,
+      "learning_rate": 0.00015,
+      "loss": 4.026426792144775,
+      "step": 2131
+    },
+    {
+      "epoch": 0.648073561820807,
+      "grad_norm": 1.982463002204895,
+      "learning_rate": 0.00015,
+      "loss": 3.7313761711120605,
+      "step": 2132
+    },
+    {
+      "epoch": 0.6483775362869519,
+      "grad_norm": 2.323707103729248,
+      "learning_rate": 0.00015,
+      "loss": 3.6761667728424072,
+      "step": 2133
+    },
+    {
+      "epoch": 0.6486815107530968,
+      "grad_norm": 2.39660906791687,
+      "learning_rate": 0.00015,
+      "loss": 3.591019630432129,
+      "step": 2134
+    },
+    {
+      "epoch": 0.6489854852192416,
+      "grad_norm": 2.2014832496643066,
+      "learning_rate": 0.00015,
+      "loss": 3.871016502380371,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6492894596853864,
+      "grad_norm": 2.111053943634033,
+      "learning_rate": 0.00015,
+      "loss": 3.936141014099121,
+      "step": 2136
+    },
+    {
+      "epoch": 0.6495934341515313,
+      "grad_norm": 1.9703108072280884,
+      "learning_rate": 0.00015,
+      "loss": 3.9203603267669678,
+      "step": 2137
+    },
+    {
+      "epoch": 0.6498974086176761,
+      "grad_norm": 2.116358518600464,
+      "learning_rate": 0.00015,
+      "loss": 4.039840221405029,
+      "step": 2138
+    },
+    {
+      "epoch": 0.6502013830838209,
+      "grad_norm": 1.9609535932540894,
+      "learning_rate": 0.00015,
+      "loss": 3.9462080001831055,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6505053575499659,
+      "grad_norm": 2.1039302349090576,
+      "learning_rate": 0.00015,
+      "loss": 3.9341206550598145,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6508093320161107,
+      "grad_norm": 2.037930965423584,
+      "learning_rate": 0.00015,
+      "loss": 3.7598373889923096,
+      "step": 2141
+    },
+    {
+      "epoch": 0.6511133064822555,
+      "grad_norm": 2.152848482131958,
+      "learning_rate": 0.00015,
+      "loss": 3.838131904602051,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6514172809484003,
+      "grad_norm": 1.573032021522522,
+      "learning_rate": 0.00015,
+      "loss": 4.055294990539551,
+      "step": 2143
+    },
+    {
+      "epoch": 0.6517212554145452,
+      "grad_norm": 3.4032397270202637,
+      "learning_rate": 0.00015,
+      "loss": 4.393263816833496,
+      "step": 2144
+    },
+    {
+      "epoch": 0.65202522988069,
+      "grad_norm": 1.9134106636047363,
+      "learning_rate": 0.00015,
+      "loss": 4.024962425231934,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6523292043468348,
+      "grad_norm": 2.1015641689300537,
+      "learning_rate": 0.00015,
+      "loss": 3.930861473083496,
+      "step": 2146
+    },
+    {
+      "epoch": 0.6526331788129797,
+      "grad_norm": 2.1650867462158203,
+      "learning_rate": 0.00015,
+      "loss": 4.300280570983887,
+      "step": 2147
+    },
+    {
+      "epoch": 0.6529371532791246,
+      "grad_norm": 2.221924304962158,
+      "learning_rate": 0.00015,
+      "loss": 3.4569196701049805,
+      "step": 2148
+    },
+    {
+      "epoch": 0.6532411277452694,
+      "grad_norm": 2.36156964302063,
+      "learning_rate": 0.00015,
+      "loss": 3.946861743927002,
+      "step": 2149
+    },
+    {
+      "epoch": 0.6535451022114143,
+      "grad_norm": 1.9368547201156616,
+      "learning_rate": 0.00015,
+      "loss": 4.057029724121094,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6538490766775591,
+      "grad_norm": 2.3797805309295654,
+      "learning_rate": 0.00015,
+      "loss": 4.022902488708496,
+      "step": 2151
+    },
+    {
+      "epoch": 0.6541530511437039,
+      "grad_norm": 2.1438372135162354,
+      "learning_rate": 0.00015,
+      "loss": 3.927823781967163,
+      "step": 2152
+    },
+    {
+      "epoch": 0.6544570256098488,
+      "grad_norm": 1.8878881931304932,
+      "learning_rate": 0.00015,
+      "loss": 4.234925270080566,
+      "step": 2153
+    },
+    {
+      "epoch": 0.6547610000759936,
+      "grad_norm": 2.615243911743164,
+      "learning_rate": 0.00015,
+      "loss": 3.634878396987915,
+      "step": 2154
+    },
+    {
+      "epoch": 0.6550649745421384,
+      "grad_norm": 2.1991312503814697,
+      "learning_rate": 0.00015,
+      "loss": 4.102030277252197,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6553689490082834,
+      "grad_norm": 1.707837462425232,
+      "learning_rate": 0.00015,
+      "loss": 3.904210090637207,
+      "step": 2156
+    },
+    {
+      "epoch": 0.6556729234744282,
+      "grad_norm": 2.0705063343048096,
+      "learning_rate": 0.00015,
+      "loss": 3.892725706100464,
+      "step": 2157
+    },
+    {
+      "epoch": 0.655976897940573,
+      "grad_norm": 2.142409324645996,
+      "learning_rate": 0.00015,
+      "loss": 3.584454298019409,
+      "step": 2158
+    },
+    {
+      "epoch": 0.6562808724067178,
+      "grad_norm": 1.8219164609909058,
+      "learning_rate": 0.00015,
+      "loss": 3.951308012008667,
+      "step": 2159
+    },
+    {
+      "epoch": 0.6565848468728627,
+      "grad_norm": 1.8940702676773071,
+      "learning_rate": 0.00015,
+      "loss": 3.729149580001831,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6568888213390075,
+      "grad_norm": 1.8623270988464355,
+      "learning_rate": 0.00015,
+      "loss": 3.8560433387756348,
+      "step": 2161
+    },
+    {
+      "epoch": 0.6571927958051523,
+      "grad_norm": 1.9557658433914185,
+      "learning_rate": 0.00015,
+      "loss": 3.9077131748199463,
+      "step": 2162
+    },
+    {
+      "epoch": 0.6574967702712972,
+      "grad_norm": 1.8589352369308472,
+      "learning_rate": 0.00015,
+      "loss": 3.9425952434539795,
+      "step": 2163
+    },
+    {
+      "epoch": 0.6578007447374421,
+      "grad_norm": 2.8029227256774902,
+      "learning_rate": 0.00015,
+      "loss": 4.167500972747803,
+      "step": 2164
+    },
+    {
+      "epoch": 0.6581047192035869,
+      "grad_norm": 2.816310167312622,
+      "learning_rate": 0.00015,
+      "loss": 4.397432804107666,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6584086936697318,
+      "grad_norm": 2.8269307613372803,
+      "learning_rate": 0.00015,
+      "loss": 3.7244536876678467,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6587126681358766,
+      "grad_norm": 1.9573314189910889,
+      "learning_rate": 0.00015,
+      "loss": 3.9227495193481445,
+      "step": 2167
+    },
+    {
+      "epoch": 0.6590166426020214,
+      "grad_norm": 1.8966187238693237,
+      "learning_rate": 0.00015,
+      "loss": 3.9842376708984375,
+      "step": 2168
+    },
+    {
+      "epoch": 0.6593206170681662,
+      "grad_norm": 2.064720630645752,
+      "learning_rate": 0.00015,
+      "loss": 3.769381523132324,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6596245915343111,
+      "grad_norm": 2.791022539138794,
+      "learning_rate": 0.00015,
+      "loss": 3.8914918899536133,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6599285660004559,
+      "grad_norm": 2.4794161319732666,
+      "learning_rate": 0.00015,
+      "loss": 4.045576095581055,
+      "step": 2171
+    },
+    {
+      "epoch": 0.6602325404666008,
+      "grad_norm": 2.127053737640381,
+      "learning_rate": 0.00015,
+      "loss": 3.951082229614258,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6605365149327457,
+      "grad_norm": 2.0484800338745117,
+      "learning_rate": 0.00015,
+      "loss": 3.8932716846466064,
+      "step": 2173
+    },
+    {
+      "epoch": 0.6608404893988905,
+      "grad_norm": 2.081547260284424,
+      "learning_rate": 0.00015,
+      "loss": 3.5715348720550537,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6611444638650353,
+      "grad_norm": 2.274630069732666,
+      "learning_rate": 0.00015,
+      "loss": 3.9247937202453613,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6614484383311802,
+      "grad_norm": 2.0000014305114746,
+      "learning_rate": 0.00015,
+      "loss": 3.7380659580230713,
+      "step": 2176
+    },
+    {
+      "epoch": 0.661752412797325,
+      "grad_norm": 1.953714370727539,
+      "learning_rate": 0.00015,
+      "loss": 3.803218364715576,
+      "step": 2177
+    },
+    {
+      "epoch": 0.6620563872634698,
+      "grad_norm": 1.887970209121704,
+      "learning_rate": 0.00015,
+      "loss": 3.7834956645965576,
+      "step": 2178
+    },
+    {
+      "epoch": 0.6623603617296148,
+      "grad_norm": 1.9931479692459106,
+      "learning_rate": 0.00015,
+      "loss": 4.014900207519531,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6626643361957596,
+      "grad_norm": 2.8677051067352295,
+      "learning_rate": 0.00015,
+      "loss": 3.6450440883636475,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6629683106619044,
+      "grad_norm": 2.0814669132232666,
+      "learning_rate": 0.00015,
+      "loss": 3.484046220779419,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6632722851280493,
+      "grad_norm": 1.831112265586853,
+      "learning_rate": 0.00015,
+      "loss": 3.941951274871826,
+      "step": 2182
+    },
+    {
+      "epoch": 0.6635762595941941,
+      "grad_norm": 1.9422144889831543,
+      "learning_rate": 0.00015,
+      "loss": 3.883584976196289,
+      "step": 2183
+    },
+    {
+      "epoch": 0.6638802340603389,
+      "grad_norm": 2.443401336669922,
+      "learning_rate": 0.00015,
+      "loss": 3.977149724960327,
+      "step": 2184
+    },
+    {
+      "epoch": 0.6641842085264837,
+      "grad_norm": 2.142738103866577,
+      "learning_rate": 0.00015,
+      "loss": 3.88832426071167,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6644881829926286,
+      "grad_norm": 1.8268678188323975,
+      "learning_rate": 0.00015,
+      "loss": 4.1069016456604,
+      "step": 2186
+    },
+    {
+      "epoch": 0.6647921574587735,
+      "grad_norm": 1.8886451721191406,
+      "learning_rate": 0.00015,
+      "loss": 3.836519718170166,
+      "step": 2187
+    },
+    {
+      "epoch": 0.6650961319249183,
+      "grad_norm": 1.9812830686569214,
+      "learning_rate": 0.00015,
+      "loss": 3.9749274253845215,
+      "step": 2188
+    },
+    {
+      "epoch": 0.6654001063910632,
+      "grad_norm": 1.8550376892089844,
+      "learning_rate": 0.00015,
+      "loss": 3.7828900814056396,
+      "step": 2189
+    },
+    {
+      "epoch": 0.665704080857208,
+      "grad_norm": 2.03593111038208,
+      "learning_rate": 0.00015,
+      "loss": 3.9889116287231445,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6660080553233528,
+      "grad_norm": 2.7390570640563965,
+      "learning_rate": 0.00015,
+      "loss": 3.692859172821045,
+      "step": 2191
+    },
+    {
+      "epoch": 0.6663120297894977,
+      "grad_norm": 2.1565961837768555,
+      "learning_rate": 0.00015,
+      "loss": 3.7681264877319336,
+      "step": 2192
+    },
+    {
+      "epoch": 0.6666160042556425,
+      "grad_norm": 1.7327367067337036,
+      "learning_rate": 0.00015,
+      "loss": 3.947014808654785,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6669199787217873,
+      "grad_norm": 2.3895821571350098,
+      "learning_rate": 0.00015,
+      "loss": 3.6060473918914795,
+      "step": 2194
+    },
+    {
+      "epoch": 0.6672239531879323,
+      "grad_norm": 1.8089061975479126,
+      "learning_rate": 0.00015,
+      "loss": 4.06124210357666,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6675279276540771,
+      "grad_norm": 1.7783255577087402,
+      "learning_rate": 0.00015,
+      "loss": 3.4932050704956055,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6678319021202219,
+      "grad_norm": 2.1997475624084473,
+      "learning_rate": 0.00015,
+      "loss": 4.012927055358887,
+      "step": 2197
+    },
+    {
+      "epoch": 0.6681358765863668,
+      "grad_norm": 2.157263994216919,
+      "learning_rate": 0.00015,
+      "loss": 3.897960662841797,
+      "step": 2198
+    },
+    {
+      "epoch": 0.6684398510525116,
+      "grad_norm": 1.7488255500793457,
+      "learning_rate": 0.00015,
+      "loss": 3.9950366020202637,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6687438255186564,
+      "grad_norm": 2.0957791805267334,
+      "learning_rate": 0.00015,
+      "loss": 4.047287464141846,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6690477999848012,
+      "grad_norm": 1.749588966369629,
+      "learning_rate": 0.00015,
+      "loss": 3.7459442615509033,
+      "step": 2201
+    },
+    {
+      "epoch": 0.6693517744509461,
+      "grad_norm": 2.9807353019714355,
+      "learning_rate": 0.00015,
+      "loss": 3.699831008911133,
+      "step": 2202
+    },
+    {
+      "epoch": 0.669655748917091,
+      "grad_norm": 1.9000790119171143,
+      "learning_rate": 0.00015,
+      "loss": 3.7796630859375,
+      "step": 2203
+    },
+    {
+      "epoch": 0.6699597233832358,
+      "grad_norm": 1.7540078163146973,
+      "learning_rate": 0.00015,
+      "loss": 3.387889862060547,
+      "step": 2204
+    },
+    {
+      "epoch": 0.6702636978493807,
+      "grad_norm": 2.2168526649475098,
+      "learning_rate": 0.00015,
+      "loss": 3.5826263427734375,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6705676723155255,
+      "grad_norm": 2.4718916416168213,
+      "learning_rate": 0.00015,
+      "loss": 3.9106876850128174,
+      "step": 2206
+    },
+    {
+      "epoch": 0.6708716467816703,
+      "grad_norm": 2.0908000469207764,
+      "learning_rate": 0.00015,
+      "loss": 3.7312324047088623,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6711756212478152,
+      "grad_norm": 3.0742359161376953,
+      "learning_rate": 0.00015,
+      "loss": 4.20989990234375,
+      "step": 2208
+    },
+    {
+      "epoch": 0.67147959571396,
+      "grad_norm": 1.8382582664489746,
+      "learning_rate": 0.00015,
+      "loss": 3.6115055084228516,
+      "step": 2209
+    },
+    {
+      "epoch": 0.6717835701801048,
+      "grad_norm": 1.9609347581863403,
+      "learning_rate": 0.00015,
+      "loss": 3.7778773307800293,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6720875446462498,
+      "grad_norm": 2.085472345352173,
+      "learning_rate": 0.00015,
+      "loss": 3.7461471557617188,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6723915191123946,
+      "grad_norm": 2.337397336959839,
+      "learning_rate": 0.00015,
+      "loss": 3.8503708839416504,
+      "step": 2212
+    },
+    {
+      "epoch": 0.6726954935785394,
+      "grad_norm": 2.0556118488311768,
+      "learning_rate": 0.00015,
+      "loss": 3.852043628692627,
+      "step": 2213
+    },
+    {
+      "epoch": 0.6729994680446842,
+      "grad_norm": 1.8260571956634521,
+      "learning_rate": 0.00015,
+      "loss": 4.116177558898926,
+      "step": 2214
+    },
+    {
+      "epoch": 0.6733034425108291,
+      "grad_norm": 2.265875816345215,
+      "learning_rate": 0.00015,
+      "loss": 3.8143815994262695,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6736074169769739,
+      "grad_norm": 2.061357021331787,
+      "learning_rate": 0.00015,
+      "loss": 3.8354060649871826,
+      "step": 2216
+    },
+    {
+      "epoch": 0.6739113914431187,
+      "grad_norm": 1.9409353733062744,
+      "learning_rate": 0.00015,
+      "loss": 4.149125099182129,
+      "step": 2217
+    },
+    {
+      "epoch": 0.6742153659092637,
+      "grad_norm": 2.007194995880127,
+      "learning_rate": 0.00015,
+      "loss": 3.7001678943634033,
+      "step": 2218
+    },
+    {
+      "epoch": 0.6745193403754085,
+      "grad_norm": 1.9448699951171875,
+      "learning_rate": 0.00015,
+      "loss": 3.7710623741149902,
+      "step": 2219
+    },
+    {
+      "epoch": 0.6748233148415533,
+      "grad_norm": 1.9094657897949219,
+      "learning_rate": 0.00015,
+      "loss": 3.5140674114227295,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6751272893076982,
+      "grad_norm": 2.1284799575805664,
+      "learning_rate": 0.00015,
+      "loss": 4.054790496826172,
+      "step": 2221
+    },
+    {
+      "epoch": 0.675431263773843,
+      "grad_norm": 2.2194747924804688,
+      "learning_rate": 0.00015,
+      "loss": 4.052712440490723,
+      "step": 2222
+    },
+    {
+      "epoch": 0.6757352382399878,
+      "grad_norm": 2.643186569213867,
+      "learning_rate": 0.00015,
+      "loss": 3.8196897506713867,
+      "step": 2223
+    },
+    {
+      "epoch": 0.6760392127061327,
+      "grad_norm": 2.4393186569213867,
+      "learning_rate": 0.00015,
+      "loss": 3.539423704147339,
+      "step": 2224
+    },
+    {
+      "epoch": 0.6763431871722775,
+      "grad_norm": 2.0636255741119385,
+      "learning_rate": 0.00015,
+      "loss": 4.036905288696289,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6766471616384224,
+      "grad_norm": 2.4061381816864014,
+      "learning_rate": 0.00015,
+      "loss": 4.19520378112793,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6769511361045673,
+      "grad_norm": 1.5326567888259888,
+      "learning_rate": 0.00015,
+      "loss": 3.7563257217407227,
+      "step": 2227
+    },
+    {
+      "epoch": 0.6772551105707121,
+      "grad_norm": 2.2298922538757324,
+      "learning_rate": 0.00015,
+      "loss": 3.4945881366729736,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6775590850368569,
+      "grad_norm": 1.70673668384552,
+      "learning_rate": 0.00015,
+      "loss": 3.915506601333618,
+      "step": 2229
+    },
+    {
+      "epoch": 0.6778630595030017,
+      "grad_norm": 2.0627079010009766,
+      "learning_rate": 0.00015,
+      "loss": 3.7202982902526855,
+      "step": 2230
+    },
+    {
+      "epoch": 0.6781670339691466,
+      "grad_norm": 2.462486505508423,
+      "learning_rate": 0.00015,
+      "loss": 3.8206980228424072,
+      "step": 2231
+    },
+    {
+      "epoch": 0.6784710084352914,
+      "grad_norm": 1.8998327255249023,
+      "learning_rate": 0.00015,
+      "loss": 4.008999824523926,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6787749829014362,
+      "grad_norm": 2.3937129974365234,
+      "learning_rate": 0.00015,
+      "loss": 3.4744443893432617,
+      "step": 2233
+    },
+    {
+      "epoch": 0.6790789573675812,
+      "grad_norm": 1.9022068977355957,
+      "learning_rate": 0.00015,
+      "loss": 3.597585678100586,
+      "step": 2234
+    },
+    {
+      "epoch": 0.679382931833726,
+      "grad_norm": 1.9682555198669434,
+      "learning_rate": 0.00015,
+      "loss": 3.972428321838379,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6796869062998708,
+      "grad_norm": 2.114656686782837,
+      "learning_rate": 0.00015,
+      "loss": 3.87371563911438,
+      "step": 2236
+    },
+    {
+      "epoch": 0.6799908807660157,
+      "grad_norm": 1.9050331115722656,
+      "learning_rate": 0.00015,
+      "loss": 3.9147393703460693,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6802948552321605,
+      "grad_norm": 1.768567442893982,
+      "learning_rate": 0.00015,
+      "loss": 3.8335323333740234,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6805988296983053,
+      "grad_norm": 2.1438584327697754,
+      "learning_rate": 0.00015,
+      "loss": 3.873220920562744,
+      "step": 2239
+    },
+    {
+      "epoch": 0.6809028041644501,
+      "grad_norm": 2.217785120010376,
+      "learning_rate": 0.00015,
+      "loss": 3.960599899291992,
+      "step": 2240
+    },
+    {
+      "epoch": 0.681206778630595,
+      "grad_norm": 1.6726840734481812,
+      "learning_rate": 0.00015,
+      "loss": 3.914271593093872,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6815107530967399,
+      "grad_norm": 2.034036636352539,
+      "learning_rate": 0.00015,
+      "loss": 3.7344412803649902,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6818147275628847,
+      "grad_norm": 1.934544563293457,
+      "learning_rate": 0.00015,
+      "loss": 3.7108049392700195,
+      "step": 2243
+    },
+    {
+      "epoch": 0.6821187020290296,
+      "grad_norm": 2.2361507415771484,
+      "learning_rate": 0.00015,
+      "loss": 3.779963970184326,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6824226764951744,
+      "grad_norm": 2.1383793354034424,
+      "learning_rate": 0.00015,
+      "loss": 3.5640339851379395,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6827266509613192,
+      "grad_norm": 2.2309558391571045,
+      "learning_rate": 0.00015,
+      "loss": 3.7917020320892334,
+      "step": 2246
+    },
+    {
+      "epoch": 0.6830306254274641,
+      "grad_norm": 1.7698349952697754,
+      "learning_rate": 0.00015,
+      "loss": 3.850095272064209,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6833345998936089,
+      "grad_norm": 1.8090314865112305,
+      "learning_rate": 0.00015,
+      "loss": 3.9251699447631836,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6836385743597537,
+      "grad_norm": 2.7806334495544434,
+      "learning_rate": 0.00015,
+      "loss": 3.8011012077331543,
+      "step": 2249
+    },
+    {
+      "epoch": 0.6839425488258987,
+      "grad_norm": 1.9666911363601685,
+      "learning_rate": 0.00015,
+      "loss": 3.9352240562438965,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6842465232920435,
+      "grad_norm": 2.006467819213867,
+      "learning_rate": 0.00015,
+      "loss": 3.703120708465576,
+      "step": 2251
+    },
+    {
+      "epoch": 0.6845504977581883,
+      "grad_norm": 2.60636043548584,
+      "learning_rate": 0.00015,
+      "loss": 3.9508275985717773,
+      "step": 2252
+    },
+    {
+      "epoch": 0.6848544722243332,
+      "grad_norm": 2.2055256366729736,
+      "learning_rate": 0.00015,
+      "loss": 4.269290924072266,
+      "step": 2253
+    },
+    {
+      "epoch": 0.685158446690478,
+      "grad_norm": 1.9666045904159546,
+      "learning_rate": 0.00015,
+      "loss": 3.9848289489746094,
+      "step": 2254
+    },
+    {
+      "epoch": 0.6854624211566228,
+      "grad_norm": 1.8579381704330444,
+      "learning_rate": 0.00015,
+      "loss": 3.9202404022216797,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6857663956227676,
+      "grad_norm": 1.86093008518219,
+      "learning_rate": 0.00015,
+      "loss": 3.70344877243042,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6860703700889126,
+      "grad_norm": 3.143125295639038,
+      "learning_rate": 0.00015,
+      "loss": 3.770192861557007,
+      "step": 2257
+    },
+    {
+      "epoch": 0.6863743445550574,
+      "grad_norm": 2.7873196601867676,
+      "learning_rate": 0.00015,
+      "loss": 3.7127556800842285,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6866783190212022,
+      "grad_norm": 1.884374976158142,
+      "learning_rate": 0.00015,
+      "loss": 3.798194408416748,
+      "step": 2259
+    },
+    {
+      "epoch": 0.6869822934873471,
+      "grad_norm": 2.015958786010742,
+      "learning_rate": 0.00015,
+      "loss": 3.797424077987671,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6872862679534919,
+      "grad_norm": 1.8221850395202637,
+      "learning_rate": 0.00015,
+      "loss": 3.847198486328125,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6875902424196367,
+      "grad_norm": 1.614977478981018,
+      "learning_rate": 0.00015,
+      "loss": 3.6238765716552734,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6878942168857816,
+      "grad_norm": 2.304469108581543,
+      "learning_rate": 0.00015,
+      "loss": 3.7098090648651123,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6881981913519264,
+      "grad_norm": 2.179147243499756,
+      "learning_rate": 0.00015,
+      "loss": 3.730653762817383,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6885021658180713,
+      "grad_norm": 2.1131560802459717,
+      "learning_rate": 0.00015,
+      "loss": 3.753246307373047,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6888061402842162,
+      "grad_norm": 1.8474833965301514,
+      "learning_rate": 0.00015,
+      "loss": 3.703768253326416,
+      "step": 2266
+    },
+    {
+      "epoch": 0.689110114750361,
+      "grad_norm": 2.3927390575408936,
+      "learning_rate": 0.00015,
+      "loss": 3.855844497680664,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6894140892165058,
+      "grad_norm": 1.9300742149353027,
+      "learning_rate": 0.00015,
+      "loss": 3.6988229751586914,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6897180636826507,
+      "grad_norm": 2.0350427627563477,
+      "learning_rate": 0.00015,
+      "loss": 3.886946201324463,
+      "step": 2269
+    },
+    {
+      "epoch": 0.6900220381487955,
+      "grad_norm": 2.021387815475464,
+      "learning_rate": 0.00015,
+      "loss": 3.843048095703125,
+      "step": 2270
+    },
+    {
+      "epoch": 0.6903260126149403,
+      "grad_norm": 2.1163620948791504,
+      "learning_rate": 0.00015,
+      "loss": 3.8807294368743896,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6906299870810851,
+      "grad_norm": 2.0594124794006348,
+      "learning_rate": 0.00015,
+      "loss": 3.8512983322143555,
+      "step": 2272
+    },
+    {
+      "epoch": 0.6909339615472301,
+      "grad_norm": 1.8502436876296997,
+      "learning_rate": 0.00015,
+      "loss": 3.8193554878234863,
+      "step": 2273
+    },
+    {
+      "epoch": 0.6912379360133749,
+      "grad_norm": 2.025238037109375,
+      "learning_rate": 0.00015,
+      "loss": 3.4880011081695557,
+      "step": 2274
+    },
+    {
+      "epoch": 0.6915419104795197,
+      "grad_norm": 1.9096338748931885,
+      "learning_rate": 0.00015,
+      "loss": 3.8136520385742188,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6918458849456646,
+      "grad_norm": 2.2336349487304688,
+      "learning_rate": 0.00015,
+      "loss": 4.245523929595947,
+      "step": 2276
+    },
+    {
+      "epoch": 0.6921498594118094,
+      "grad_norm": 2.0304243564605713,
+      "learning_rate": 0.00015,
+      "loss": 4.045460224151611,
+      "step": 2277
+    },
+    {
+      "epoch": 0.6924538338779542,
+      "grad_norm": 2.382624387741089,
+      "learning_rate": 0.00015,
+      "loss": 3.757981061935425,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6927578083440991,
+      "grad_norm": 2.061579942703247,
+      "learning_rate": 0.00015,
+      "loss": 3.5691590309143066,
+      "step": 2279
+    },
+    {
+      "epoch": 0.6930617828102439,
+      "grad_norm": 2.8436734676361084,
+      "learning_rate": 0.00015,
+      "loss": 3.8829445838928223,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6933657572763888,
+      "grad_norm": 1.5905228853225708,
+      "learning_rate": 0.00015,
+      "loss": 3.5336689949035645,
+      "step": 2281
+    },
+    {
+      "epoch": 0.6936697317425337,
+      "grad_norm": 2.049947500228882,
+      "learning_rate": 0.00015,
+      "loss": 3.70709228515625,
+      "step": 2282
+    },
+    {
+      "epoch": 0.6939737062086785,
+      "grad_norm": 1.5257654190063477,
+      "learning_rate": 0.00015,
+      "loss": 3.819274425506592,
+      "step": 2283
+    },
+    {
+      "epoch": 0.6942776806748233,
+      "grad_norm": 2.6608409881591797,
+      "learning_rate": 0.00015,
+      "loss": 4.057458877563477,
+      "step": 2284
+    },
+    {
+      "epoch": 0.6945816551409681,
+      "grad_norm": 1.903224229812622,
+      "learning_rate": 0.00015,
+      "loss": 3.9113337993621826,
+      "step": 2285
+    },
+    {
+      "epoch": 0.694885629607113,
+      "grad_norm": 1.783656358718872,
+      "learning_rate": 0.00015,
+      "loss": 4.009866237640381,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6951896040732578,
+      "grad_norm": 1.9959403276443481,
+      "learning_rate": 0.00015,
+      "loss": 3.7082467079162598,
+      "step": 2287
+    },
+    {
+      "epoch": 0.6954935785394026,
+      "grad_norm": 2.0319082736968994,
+      "learning_rate": 0.00015,
+      "loss": 3.6149086952209473,
+      "step": 2288
+    },
+    {
+      "epoch": 0.6957975530055476,
+      "grad_norm": 1.8257591724395752,
+      "learning_rate": 0.00015,
+      "loss": 3.6093668937683105,
+      "step": 2289
+    },
+    {
+      "epoch": 0.6961015274716924,
+      "grad_norm": 2.0273172855377197,
+      "learning_rate": 0.00015,
+      "loss": 3.865910053253174,
+      "step": 2290
+    },
+    {
+      "epoch": 0.6964055019378372,
+      "grad_norm": 2.822133779525757,
+      "learning_rate": 0.00015,
+      "loss": 3.629441738128662,
+      "step": 2291
+    },
+    {
+      "epoch": 0.6967094764039821,
+      "grad_norm": 2.2576310634613037,
+      "learning_rate": 0.00015,
+      "loss": 3.8015079498291016,
+      "step": 2292
+    },
+    {
+      "epoch": 0.6970134508701269,
+      "grad_norm": 2.0102345943450928,
+      "learning_rate": 0.00015,
+      "loss": 3.770796537399292,
+      "step": 2293
+    },
+    {
+      "epoch": 0.6973174253362717,
+      "grad_norm": 1.8922706842422485,
+      "learning_rate": 0.00015,
+      "loss": 4.2164411544799805,
+      "step": 2294
+    },
+    {
+      "epoch": 0.6976213998024166,
+      "grad_norm": 3.371467351913452,
+      "learning_rate": 0.00015,
+      "loss": 3.9799387454986572,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6979253742685615,
+      "grad_norm": 2.412734031677246,
+      "learning_rate": 0.00015,
+      "loss": 4.000723838806152,
+      "step": 2296
+    },
+    {
+      "epoch": 0.6982293487347063,
+      "grad_norm": 2.8178484439849854,
+      "learning_rate": 0.00015,
+      "loss": 3.9026143550872803,
+      "step": 2297
+    },
+    {
+      "epoch": 0.6985333232008512,
+      "grad_norm": 1.9167522192001343,
+      "learning_rate": 0.00015,
+      "loss": 3.8683876991271973,
+      "step": 2298
+    },
+    {
+      "epoch": 0.698837297666996,
+      "grad_norm": 2.345639228820801,
+      "learning_rate": 0.00015,
+      "loss": 4.010730743408203,
+      "step": 2299
+    },
+    {
+      "epoch": 0.6991412721331408,
+      "grad_norm": 2.3848888874053955,
+      "learning_rate": 0.00015,
+      "loss": 3.7679126262664795,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6994452465992856,
+      "grad_norm": 2.0483319759368896,
+      "learning_rate": 0.00015,
+      "loss": 3.7197108268737793,
+      "step": 2301
+    },
+    {
+      "epoch": 0.6997492210654305,
+      "grad_norm": 2.696037530899048,
+      "learning_rate": 0.00015,
+      "loss": 4.009227752685547,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7000531955315753,
+      "grad_norm": 2.6747019290924072,
+      "learning_rate": 0.00015,
+      "loss": 3.5688250064849854,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7003571699977202,
+      "grad_norm": 1.987487554550171,
+      "learning_rate": 0.00015,
+      "loss": 3.5815677642822266,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7006611444638651,
+      "grad_norm": 2.004199504852295,
+      "learning_rate": 0.00015,
+      "loss": 3.9531571865081787,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7009651189300099,
+      "grad_norm": 1.7914382219314575,
+      "learning_rate": 0.00015,
+      "loss": 3.8996458053588867,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7012690933961547,
+      "grad_norm": 2.314868927001953,
+      "learning_rate": 0.00015,
+      "loss": 3.7758913040161133,
+      "step": 2307
+    },
+    {
+      "epoch": 0.7015730678622996,
+      "grad_norm": 1.6020677089691162,
+      "learning_rate": 0.00015,
+      "loss": 3.8749148845672607,
+      "step": 2308
+    },
+    {
+      "epoch": 0.7018770423284444,
+      "grad_norm": 2.0470988750457764,
+      "learning_rate": 0.00015,
+      "loss": 3.7848517894744873,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7021810167945892,
+      "grad_norm": 2.2647218704223633,
+      "learning_rate": 0.00015,
+      "loss": 3.9263758659362793,
+      "step": 2310
+    },
+    {
+      "epoch": 0.702484991260734,
+      "grad_norm": 1.8974549770355225,
+      "learning_rate": 0.00015,
+      "loss": 3.928311347961426,
+      "step": 2311
+    },
+    {
+      "epoch": 0.702788965726879,
+      "grad_norm": 1.8893988132476807,
+      "learning_rate": 0.00015,
+      "loss": 4.007492542266846,
+      "step": 2312
+    },
+    {
+      "epoch": 0.7030929401930238,
+      "grad_norm": 2.0810024738311768,
+      "learning_rate": 0.00015,
+      "loss": 3.714240550994873,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7033969146591686,
+      "grad_norm": 1.9803075790405273,
+      "learning_rate": 0.00015,
+      "loss": 4.209392070770264,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7037008891253135,
+      "grad_norm": 2.069950819015503,
+      "learning_rate": 0.00015,
+      "loss": 4.165000915527344,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7040048635914583,
+      "grad_norm": 2.1842126846313477,
+      "learning_rate": 0.00015,
+      "loss": 3.8059773445129395,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7043088380576031,
+      "grad_norm": 1.8348190784454346,
+      "learning_rate": 0.00015,
+      "loss": 3.9974029064178467,
+      "step": 2317
+    },
+    {
+      "epoch": 0.704612812523748,
+      "grad_norm": 1.928904414176941,
+      "learning_rate": 0.00015,
+      "loss": 3.796438217163086,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7049167869898928,
+      "grad_norm": 1.6367883682250977,
+      "learning_rate": 0.00015,
+      "loss": 3.9050545692443848,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7052207614560377,
+      "grad_norm": 1.7297779321670532,
+      "learning_rate": 0.00015,
+      "loss": 3.8622844219207764,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7055247359221826,
+      "grad_norm": 2.2634334564208984,
+      "learning_rate": 0.00015,
+      "loss": 4.195840835571289,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7058287103883274,
+      "grad_norm": 2.153397798538208,
+      "learning_rate": 0.00015,
+      "loss": 4.09432315826416,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7061326848544722,
+      "grad_norm": 1.9891775846481323,
+      "learning_rate": 0.00015,
+      "loss": 3.6152851581573486,
+      "step": 2323
+    },
+    {
+      "epoch": 0.706436659320617,
+      "grad_norm": 2.4357810020446777,
+      "learning_rate": 0.00015,
+      "loss": 3.783275604248047,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7067406337867619,
+      "grad_norm": 1.998834490776062,
+      "learning_rate": 0.00015,
+      "loss": 3.7565865516662598,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7070446082529067,
+      "grad_norm": 1.798938274383545,
+      "learning_rate": 0.00015,
+      "loss": 4.134984016418457,
+      "step": 2326
+    },
+    {
+      "epoch": 0.7073485827190517,
+      "grad_norm": 2.000609874725342,
+      "learning_rate": 0.00015,
+      "loss": 3.71829891204834,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7076525571851965,
+      "grad_norm": 2.3191003799438477,
+      "learning_rate": 0.00015,
+      "loss": 3.9228224754333496,
+      "step": 2328
+    },
+    {
+      "epoch": 0.7079565316513413,
+      "grad_norm": 2.140761375427246,
+      "learning_rate": 0.00015,
+      "loss": 3.9250001907348633,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7082605061174861,
+      "grad_norm": 1.7670974731445312,
+      "learning_rate": 0.00015,
+      "loss": 3.870420217514038,
+      "step": 2330
+    },
+    {
+      "epoch": 0.708564480583631,
+      "grad_norm": 1.8694233894348145,
+      "learning_rate": 0.00015,
+      "loss": 4.007411479949951,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7088684550497758,
+      "grad_norm": 1.9256595373153687,
+      "learning_rate": 0.00015,
+      "loss": 3.9210948944091797,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7091724295159206,
+      "grad_norm": 1.8244167566299438,
+      "learning_rate": 0.00015,
+      "loss": 3.6683640480041504,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7094764039820655,
+      "grad_norm": 2.0211739540100098,
+      "learning_rate": 0.00015,
+      "loss": 3.623746633529663,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7097803784482104,
+      "grad_norm": 1.869234561920166,
+      "learning_rate": 0.00015,
+      "loss": 4.060471057891846,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7100843529143552,
+      "grad_norm": 2.503586769104004,
+      "learning_rate": 0.00015,
+      "loss": 3.788425922393799,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7103883273805001,
+      "grad_norm": 2.538632392883301,
+      "learning_rate": 0.00015,
+      "loss": 4.317087173461914,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7106923018466449,
+      "grad_norm": 2.0705387592315674,
+      "learning_rate": 0.00015,
+      "loss": 3.6653623580932617,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7109962763127897,
+      "grad_norm": 1.7053858041763306,
+      "learning_rate": 0.00015,
+      "loss": 4.097198963165283,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7113002507789346,
+      "grad_norm": 1.7496962547302246,
+      "learning_rate": 0.00015,
+      "loss": 3.636474609375,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7116042252450794,
+      "grad_norm": 1.9964141845703125,
+      "learning_rate": 0.00015,
+      "loss": 3.790966510772705,
+      "step": 2341
+    },
+    {
+      "epoch": 0.7119081997112242,
+      "grad_norm": 1.6958880424499512,
+      "learning_rate": 0.00015,
+      "loss": 3.820672035217285,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7122121741773692,
+      "grad_norm": 1.7458387613296509,
+      "learning_rate": 0.00015,
+      "loss": 3.7400271892547607,
+      "step": 2343
+    },
+    {
+      "epoch": 0.712516148643514,
+      "grad_norm": 1.8991854190826416,
+      "learning_rate": 0.00015,
+      "loss": 3.699939727783203,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7128201231096588,
+      "grad_norm": 1.9006024599075317,
+      "learning_rate": 0.00015,
+      "loss": 3.6776747703552246,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7131240975758036,
+      "grad_norm": 1.7334448099136353,
+      "learning_rate": 0.00015,
+      "loss": 3.9248104095458984,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7134280720419485,
+      "grad_norm": 2.486135959625244,
+      "learning_rate": 0.00015,
+      "loss": 3.5465826988220215,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7137320465080933,
+      "grad_norm": 2.095580577850342,
+      "learning_rate": 0.00015,
+      "loss": 3.8855197429656982,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7140360209742381,
+      "grad_norm": 2.59340763092041,
+      "learning_rate": 0.00015,
+      "loss": 3.7139482498168945,
+      "step": 2349
+    },
+    {
+      "epoch": 0.714339995440383,
+      "grad_norm": 2.0470190048217773,
+      "learning_rate": 0.00015,
+      "loss": 3.9708454608917236,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7146439699065279,
+      "grad_norm": 2.0301835536956787,
+      "learning_rate": 0.00015,
+      "loss": 3.831225872039795,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7149479443726727,
+      "grad_norm": 2.2100558280944824,
+      "learning_rate": 0.00015,
+      "loss": 3.4555375576019287,
+      "step": 2352
+    },
+    {
+      "epoch": 0.7152519188388176,
+      "grad_norm": 2.4985127449035645,
+      "learning_rate": 0.00015,
+      "loss": 3.6311168670654297,
+      "step": 2353
+    },
+    {
+      "epoch": 0.7155558933049624,
+      "grad_norm": 1.9503060579299927,
+      "learning_rate": 0.00015,
+      "loss": 3.5727555751800537,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7158598677711072,
+      "grad_norm": 2.0909805297851562,
+      "learning_rate": 0.00015,
+      "loss": 3.8778200149536133,
+      "step": 2355
+    },
+    {
+      "epoch": 0.716163842237252,
+      "grad_norm": 3.1643714904785156,
+      "learning_rate": 0.00015,
+      "loss": 3.939124345779419,
+      "step": 2356
+    },
+    {
+      "epoch": 0.7164678167033969,
+      "grad_norm": 1.7228221893310547,
+      "learning_rate": 0.00015,
+      "loss": 3.4679346084594727,
+      "step": 2357
+    },
+    {
+      "epoch": 0.7167717911695417,
+      "grad_norm": 2.1850850582122803,
+      "learning_rate": 0.00015,
+      "loss": 3.8333323001861572,
+      "step": 2358
+    },
+    {
+      "epoch": 0.7170757656356866,
+      "grad_norm": 2.1210079193115234,
+      "learning_rate": 0.00015,
+      "loss": 3.5915379524230957,
+      "step": 2359
+    },
+    {
+      "epoch": 0.7173797401018315,
+      "grad_norm": 2.5463478565216064,
+      "learning_rate": 0.00015,
+      "loss": 3.8738739490509033,
+      "step": 2360
+    },
+    {
+      "epoch": 0.7176837145679763,
+      "grad_norm": 2.038787364959717,
+      "learning_rate": 0.00015,
+      "loss": 3.9073867797851562,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7179876890341211,
+      "grad_norm": 5.169443130493164,
+      "learning_rate": 0.00015,
+      "loss": 3.9399983882904053,
+      "step": 2362
+    },
+    {
+      "epoch": 0.718291663500266,
+      "grad_norm": 2.0485613346099854,
+      "learning_rate": 0.00015,
+      "loss": 3.7773396968841553,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7185956379664108,
+      "grad_norm": 2.0891642570495605,
+      "learning_rate": 0.00015,
+      "loss": 3.9943575859069824,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7188996124325556,
+      "grad_norm": 1.828840732574463,
+      "learning_rate": 0.00015,
+      "loss": 3.714021682739258,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7192035868987006,
+      "grad_norm": 1.8366401195526123,
+      "learning_rate": 0.00015,
+      "loss": 3.7821884155273438,
+      "step": 2366
+    },
+    {
+      "epoch": 0.7195075613648454,
+      "grad_norm": 1.6870390176773071,
+      "learning_rate": 0.00015,
+      "loss": 3.8114118576049805,
+      "step": 2367
+    },
+    {
+      "epoch": 0.7198115358309902,
+      "grad_norm": 2.218478202819824,
+      "learning_rate": 0.00015,
+      "loss": 3.8240575790405273,
+      "step": 2368
+    },
+    {
+      "epoch": 0.720115510297135,
+      "grad_norm": 1.9279720783233643,
+      "learning_rate": 0.00015,
+      "loss": 3.5682132244110107,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7204194847632799,
+      "grad_norm": 2.1509528160095215,
+      "learning_rate": 0.00015,
+      "loss": 3.535776138305664,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7207234592294247,
+      "grad_norm": 1.9771380424499512,
+      "learning_rate": 0.00015,
+      "loss": 3.8637657165527344,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7210274336955695,
+      "grad_norm": 1.7646435499191284,
+      "learning_rate": 0.00015,
+      "loss": 3.754662275314331,
+      "step": 2372
+    },
+    {
+      "epoch": 0.7213314081617144,
+      "grad_norm": 2.250561475753784,
+      "learning_rate": 0.00015,
+      "loss": 3.969456195831299,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7216353826278593,
+      "grad_norm": 1.7274500131607056,
+      "learning_rate": 0.00015,
+      "loss": 3.463789463043213,
+      "step": 2374
+    },
+    {
+      "epoch": 0.7219393570940041,
+      "grad_norm": 2.1791675090789795,
+      "learning_rate": 0.00015,
+      "loss": 3.737630605697632,
+      "step": 2375
+    },
+    {
+      "epoch": 0.722243331560149,
+      "grad_norm": 1.8988982439041138,
+      "learning_rate": 0.00015,
+      "loss": 3.983247995376587,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7225473060262938,
+      "grad_norm": 1.7831110954284668,
+      "learning_rate": 0.00015,
+      "loss": 3.4409332275390625,
+      "step": 2377
+    },
+    {
+      "epoch": 0.7228512804924386,
+      "grad_norm": 1.9970145225524902,
+      "learning_rate": 0.00015,
+      "loss": 3.7903366088867188,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7231552549585835,
+      "grad_norm": 1.685693383216858,
+      "learning_rate": 0.00015,
+      "loss": 3.829896926879883,
+      "step": 2379
+    },
+    {
+      "epoch": 0.7234592294247283,
+      "grad_norm": 1.7337901592254639,
+      "learning_rate": 0.00015,
+      "loss": 3.800349712371826,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7237632038908731,
+      "grad_norm": 2.3080403804779053,
+      "learning_rate": 0.00015,
+      "loss": 3.689204216003418,
+      "step": 2381
+    },
+    {
+      "epoch": 0.7240671783570181,
+      "grad_norm": 2.062032699584961,
+      "learning_rate": 0.00015,
+      "loss": 4.045444488525391,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7243711528231629,
+      "grad_norm": 2.4142401218414307,
+      "learning_rate": 0.00015,
+      "loss": 3.865565299987793,
+      "step": 2383
+    },
+    {
+      "epoch": 0.7246751272893077,
+      "grad_norm": 2.183976411819458,
+      "learning_rate": 0.00015,
+      "loss": 3.606846570968628,
+      "step": 2384
+    },
+    {
+      "epoch": 0.7249791017554525,
+      "grad_norm": 1.8973685503005981,
+      "learning_rate": 0.00015,
+      "loss": 3.85847806930542,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7252830762215974,
+      "grad_norm": 2.0936782360076904,
+      "learning_rate": 0.00015,
+      "loss": 3.6020398139953613,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7255870506877422,
+      "grad_norm": 2.1699774265289307,
+      "learning_rate": 0.00015,
+      "loss": 3.7674131393432617,
+      "step": 2387
+    },
+    {
+      "epoch": 0.725891025153887,
+      "grad_norm": 2.07738995552063,
+      "learning_rate": 0.00015,
+      "loss": 3.900813341140747,
+      "step": 2388
+    },
+    {
+      "epoch": 0.7261949996200319,
+      "grad_norm": 2.6083099842071533,
+      "learning_rate": 0.00015,
+      "loss": 3.783237934112549,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7264989740861768,
+      "grad_norm": 1.65462064743042,
+      "learning_rate": 0.00015,
+      "loss": 4.047060012817383,
+      "step": 2390
+    },
+    {
+      "epoch": 0.7268029485523216,
+      "grad_norm": 2.255364179611206,
+      "learning_rate": 0.00015,
+      "loss": 3.9777469635009766,
+      "step": 2391
+    },
+    {
+      "epoch": 0.7271069230184665,
+      "grad_norm": 1.9162055253982544,
+      "learning_rate": 0.00015,
+      "loss": 3.9974424839019775,
+      "step": 2392
+    },
+    {
+      "epoch": 0.7274108974846113,
+      "grad_norm": 2.271000385284424,
+      "learning_rate": 0.00015,
+      "loss": 3.9004621505737305,
+      "step": 2393
+    },
+    {
+      "epoch": 0.7277148719507561,
+      "grad_norm": 1.964870810508728,
+      "learning_rate": 0.00015,
+      "loss": 3.599379539489746,
+      "step": 2394
+    },
+    {
+      "epoch": 0.728018846416901,
+      "grad_norm": 1.8590707778930664,
+      "learning_rate": 0.00015,
+      "loss": 3.6936378479003906,
+      "step": 2395
+    },
+    {
+      "epoch": 0.7283228208830458,
+      "grad_norm": 1.8527753353118896,
+      "learning_rate": 0.00015,
+      "loss": 3.841982841491699,
+      "step": 2396
+    },
+    {
+      "epoch": 0.7286267953491906,
+      "grad_norm": 2.01893949508667,
+      "learning_rate": 0.00015,
+      "loss": 3.7174599170684814,
+      "step": 2397
+    },
+    {
+      "epoch": 0.7289307698153356,
+      "grad_norm": 2.0561673641204834,
+      "learning_rate": 0.00015,
+      "loss": 4.138547897338867,
+      "step": 2398
+    },
+    {
+      "epoch": 0.7292347442814804,
+      "grad_norm": 1.9961222410202026,
+      "learning_rate": 0.00015,
+      "loss": 3.639758586883545,
+      "step": 2399
+    },
+    {
+      "epoch": 0.7295387187476252,
+      "grad_norm": 1.9173557758331299,
+      "learning_rate": 0.00015,
+      "loss": 4.148599624633789,
+      "step": 2400
+    },
+    {
+      "epoch": 0.72984269321377,
+      "grad_norm": 1.7599033117294312,
+      "learning_rate": 0.00015,
+      "loss": 4.393734931945801,
+      "step": 2401
+    },
+    {
+      "epoch": 0.7301466676799149,
+      "grad_norm": 2.578235149383545,
+      "learning_rate": 0.00015,
+      "loss": 3.7964372634887695,
+      "step": 2402
+    },
+    {
+      "epoch": 0.7304506421460597,
+      "grad_norm": 2.7241225242614746,
+      "learning_rate": 0.00015,
+      "loss": 3.684929847717285,
+      "step": 2403
+    },
+    {
+      "epoch": 0.7307546166122045,
+      "grad_norm": 2.2655298709869385,
+      "learning_rate": 0.00015,
+      "loss": 3.80268931388855,
+      "step": 2404
+    },
+    {
+      "epoch": 0.7310585910783495,
+      "grad_norm": 1.8875256776809692,
+      "learning_rate": 0.00015,
+      "loss": 3.6280112266540527,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7313625655444943,
+      "grad_norm": 1.9950344562530518,
+      "learning_rate": 0.00015,
+      "loss": 3.9896657466888428,
+      "step": 2406
+    },
+    {
+      "epoch": 0.7316665400106391,
+      "grad_norm": 2.065897226333618,
+      "learning_rate": 0.00015,
+      "loss": 4.028974533081055,
+      "step": 2407
+    },
+    {
+      "epoch": 0.731970514476784,
+      "grad_norm": 2.4203271865844727,
+      "learning_rate": 0.00015,
+      "loss": 3.819841146469116,
+      "step": 2408
+    },
+    {
+      "epoch": 0.7322744889429288,
+      "grad_norm": 2.051347494125366,
+      "learning_rate": 0.00015,
+      "loss": 4.001091003417969,
+      "step": 2409
+    },
+    {
+      "epoch": 0.7325784634090736,
+      "grad_norm": 2.0566062927246094,
+      "learning_rate": 0.00015,
+      "loss": 3.728990077972412,
+      "step": 2410
+    },
+    {
+      "epoch": 0.7328824378752185,
+      "grad_norm": 1.755820631980896,
+      "learning_rate": 0.00015,
+      "loss": 4.1898417472839355,
+      "step": 2411
+    },
+    {
+      "epoch": 0.7331864123413633,
+      "grad_norm": 2.4151909351348877,
+      "learning_rate": 0.00015,
+      "loss": 3.8205695152282715,
+      "step": 2412
+    },
+    {
+      "epoch": 0.7334903868075082,
+      "grad_norm": 2.1608948707580566,
+      "learning_rate": 0.00015,
+      "loss": 3.814892053604126,
+      "step": 2413
+    },
+    {
+      "epoch": 0.733794361273653,
+      "grad_norm": 1.961190104484558,
+      "learning_rate": 0.00015,
+      "loss": 3.7875351905822754,
+      "step": 2414
+    },
+    {
+      "epoch": 0.7340983357397979,
+      "grad_norm": 1.843995213508606,
+      "learning_rate": 0.00015,
+      "loss": 3.7175679206848145,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7344023102059427,
+      "grad_norm": 1.5962507724761963,
+      "learning_rate": 0.00015,
+      "loss": 4.062472343444824,
+      "step": 2416
+    },
+    {
+      "epoch": 0.7347062846720875,
+      "grad_norm": 2.0225741863250732,
+      "learning_rate": 0.00015,
+      "loss": 3.673892021179199,
+      "step": 2417
+    },
+    {
+      "epoch": 0.7350102591382324,
+      "grad_norm": 1.861592411994934,
+      "learning_rate": 0.00015,
+      "loss": 3.7030582427978516,
+      "step": 2418
+    },
+    {
+      "epoch": 0.7353142336043772,
+      "grad_norm": 1.6833539009094238,
+      "learning_rate": 0.00015,
+      "loss": 3.952922821044922,
+      "step": 2419
+    },
+    {
+      "epoch": 0.735618208070522,
+      "grad_norm": 2.248894691467285,
+      "learning_rate": 0.00015,
+      "loss": 3.635101079940796,
+      "step": 2420
+    },
+    {
+      "epoch": 0.735922182536667,
+      "grad_norm": 3.2242486476898193,
+      "learning_rate": 0.00015,
+      "loss": 3.925743579864502,
+      "step": 2421
+    },
+    {
+      "epoch": 0.7362261570028118,
+      "grad_norm": 2.1128928661346436,
+      "learning_rate": 0.00015,
+      "loss": 3.298746347427368,
+      "step": 2422
+    },
+    {
+      "epoch": 0.7365301314689566,
+      "grad_norm": 2.4417901039123535,
+      "learning_rate": 0.00015,
+      "loss": 3.537980079650879,
+      "step": 2423
+    },
+    {
+      "epoch": 0.7368341059351015,
+      "grad_norm": 2.6412925720214844,
+      "learning_rate": 0.00015,
+      "loss": 3.8073997497558594,
+      "step": 2424
+    },
+    {
+      "epoch": 0.7371380804012463,
+      "grad_norm": 2.3333370685577393,
+      "learning_rate": 0.00015,
+      "loss": 3.8592119216918945,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7374420548673911,
+      "grad_norm": 2.431441068649292,
+      "learning_rate": 0.00015,
+      "loss": 3.5997722148895264,
+      "step": 2426
+    },
+    {
+      "epoch": 0.737746029333536,
+      "grad_norm": 1.8683111667633057,
+      "learning_rate": 0.00015,
+      "loss": 3.838843822479248,
+      "step": 2427
+    },
+    {
+      "epoch": 0.7380500037996808,
+      "grad_norm": 1.9979158639907837,
+      "learning_rate": 0.00015,
+      "loss": 3.890564441680908,
+      "step": 2428
+    },
+    {
+      "epoch": 0.7383539782658257,
+      "grad_norm": 2.0609686374664307,
+      "learning_rate": 0.00015,
+      "loss": 3.75339937210083,
+      "step": 2429
+    },
+    {
+      "epoch": 0.7386579527319705,
+      "grad_norm": 2.3087377548217773,
+      "learning_rate": 0.00015,
+      "loss": 3.519409418106079,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7389619271981154,
+      "grad_norm": 1.8774174451828003,
+      "learning_rate": 0.00015,
+      "loss": 3.7129874229431152,
+      "step": 2431
+    },
+    {
+      "epoch": 0.7392659016642602,
+      "grad_norm": 1.7575942277908325,
+      "learning_rate": 0.00015,
+      "loss": 3.463355541229248,
+      "step": 2432
+    },
+    {
+      "epoch": 0.739569876130405,
+      "grad_norm": 2.3393075466156006,
+      "learning_rate": 0.00015,
+      "loss": 3.9968950748443604,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7398738505965499,
+      "grad_norm": 2.0355591773986816,
+      "learning_rate": 0.00015,
+      "loss": 4.313406944274902,
+      "step": 2434
+    },
+    {
+      "epoch": 0.7401778250626947,
+      "grad_norm": 2.3697402477264404,
+      "learning_rate": 0.00015,
+      "loss": 3.8104982376098633,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7404817995288395,
+      "grad_norm": 1.8974552154541016,
+      "learning_rate": 0.00015,
+      "loss": 3.5748677253723145,
+      "step": 2436
+    },
+    {
+      "epoch": 0.7407857739949845,
+      "grad_norm": 1.8295938968658447,
+      "learning_rate": 0.00015,
+      "loss": 3.627169609069824,
+      "step": 2437
+    },
+    {
+      "epoch": 0.7410897484611293,
+      "grad_norm": 1.825971007347107,
+      "learning_rate": 0.00015,
+      "loss": 4.113907814025879,
+      "step": 2438
+    },
+    {
+      "epoch": 0.7413937229272741,
+      "grad_norm": 1.9154871702194214,
+      "learning_rate": 0.00015,
+      "loss": 3.8848628997802734,
+      "step": 2439
+    },
+    {
+      "epoch": 0.741697697393419,
+      "grad_norm": 1.756745457649231,
+      "learning_rate": 0.00015,
+      "loss": 3.6349902153015137,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7420016718595638,
+      "grad_norm": 1.999097466468811,
+      "learning_rate": 0.00015,
+      "loss": 3.662959098815918,
+      "step": 2441
+    },
+    {
+      "epoch": 0.7423056463257086,
+      "grad_norm": 2.0709376335144043,
+      "learning_rate": 0.00015,
+      "loss": 3.6555562019348145,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7426096207918534,
+      "grad_norm": 1.9014334678649902,
+      "learning_rate": 0.00015,
+      "loss": 3.8010592460632324,
+      "step": 2443
+    },
+    {
+      "epoch": 0.7429135952579984,
+      "grad_norm": 1.883528232574463,
+      "learning_rate": 0.00015,
+      "loss": 3.815516948699951,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7432175697241432,
+      "grad_norm": 2.4213809967041016,
+      "learning_rate": 0.00015,
+      "loss": 4.0446858406066895,
+      "step": 2445
+    },
+    {
+      "epoch": 0.743521544190288,
+      "grad_norm": 1.7499926090240479,
+      "learning_rate": 0.00015,
+      "loss": 4.001878261566162,
+      "step": 2446
+    },
+    {
+      "epoch": 0.7438255186564329,
+      "grad_norm": 1.859399437904358,
+      "learning_rate": 0.00015,
+      "loss": 3.5954794883728027,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7441294931225777,
+      "grad_norm": 2.054652452468872,
+      "learning_rate": 0.00015,
+      "loss": 3.9973909854888916,
+      "step": 2448
+    },
+    {
+      "epoch": 0.7444334675887225,
+      "grad_norm": 1.8208959102630615,
+      "learning_rate": 0.00015,
+      "loss": 3.697951316833496,
+      "step": 2449
+    },
+    {
+      "epoch": 0.7447374420548674,
+      "grad_norm": 2.023566722869873,
+      "learning_rate": 0.00015,
+      "loss": 4.138322830200195,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7450414165210122,
+      "grad_norm": 3.8194351196289062,
+      "learning_rate": 0.00015,
+      "loss": 3.993570566177368,
+      "step": 2451
+    },
+    {
+      "epoch": 0.7453453909871571,
+      "grad_norm": 2.2663469314575195,
+      "learning_rate": 0.00015,
+      "loss": 3.5742950439453125,
+      "step": 2452
+    },
+    {
+      "epoch": 0.745649365453302,
+      "grad_norm": 1.9613299369812012,
+      "learning_rate": 0.00015,
+      "loss": 3.4578890800476074,
+      "step": 2453
+    },
+    {
+      "epoch": 0.7459533399194468,
+      "grad_norm": 1.9861711263656616,
+      "learning_rate": 0.00015,
+      "loss": 3.4439449310302734,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7462573143855916,
+      "grad_norm": 2.2554337978363037,
+      "learning_rate": 0.00015,
+      "loss": 3.4629242420196533,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7465612888517364,
+      "grad_norm": 2.066869020462036,
+      "learning_rate": 0.00015,
+      "loss": 3.6736040115356445,
+      "step": 2456
+    },
+    {
+      "epoch": 0.7468652633178813,
+      "grad_norm": 1.756807804107666,
+      "learning_rate": 0.00015,
+      "loss": 3.742140769958496,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7471692377840261,
+      "grad_norm": 1.858678936958313,
+      "learning_rate": 0.00015,
+      "loss": 3.5473809242248535,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7474732122501709,
+      "grad_norm": 1.6141929626464844,
+      "learning_rate": 0.00015,
+      "loss": 4.1063995361328125,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7477771867163159,
+      "grad_norm": 4.166224479675293,
+      "learning_rate": 0.00015,
+      "loss": 3.9083774089813232,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7480811611824607,
+      "grad_norm": 2.3126049041748047,
+      "learning_rate": 0.00015,
+      "loss": 3.8495171070098877,
+      "step": 2461
+    },
+    {
+      "epoch": 0.7483851356486055,
+      "grad_norm": 2.3565471172332764,
+      "learning_rate": 0.00015,
+      "loss": 3.412517786026001,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7486891101147504,
+      "grad_norm": 2.097257137298584,
+      "learning_rate": 0.00015,
+      "loss": 3.8410439491271973,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7489930845808952,
+      "grad_norm": 1.5394903421401978,
+      "learning_rate": 0.00015,
+      "loss": 3.710925579071045,
+      "step": 2464
+    },
+    {
+      "epoch": 0.74929705904704,
+      "grad_norm": 1.9403960704803467,
+      "learning_rate": 0.00015,
+      "loss": 3.77772855758667,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7496010335131849,
+      "grad_norm": 2.0482542514801025,
+      "learning_rate": 0.00015,
+      "loss": 3.720806121826172,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7499050079793297,
+      "grad_norm": 1.9034346342086792,
+      "learning_rate": 0.00015,
+      "loss": 3.8812994956970215,
+      "step": 2467
+    },
+    {
+      "epoch": 0.7502089824454746,
+      "grad_norm": 1.8663387298583984,
+      "learning_rate": 0.00015,
+      "loss": 3.6654021739959717,
+      "step": 2468
+    },
+    {
+      "epoch": 0.7505129569116195,
+      "grad_norm": 6.920061111450195,
+      "learning_rate": 0.00015,
+      "loss": 3.5962376594543457,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7508169313777643,
+      "grad_norm": 1.676918864250183,
+      "learning_rate": 0.00015,
+      "loss": 3.7969653606414795,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7511209058439091,
+      "grad_norm": 1.9080584049224854,
+      "learning_rate": 0.00015,
+      "loss": 3.9017648696899414,
+      "step": 2471
+    },
+    {
+      "epoch": 0.7514248803100539,
+      "grad_norm": 1.8772486448287964,
+      "learning_rate": 0.00015,
+      "loss": 3.8930296897888184,
+      "step": 2472
+    },
+    {
+      "epoch": 0.7517288547761988,
+      "grad_norm": 2.249570846557617,
+      "learning_rate": 0.00015,
+      "loss": 3.857417106628418,
+      "step": 2473
+    },
+    {
+      "epoch": 0.7520328292423436,
+      "grad_norm": 2.1502034664154053,
+      "learning_rate": 0.00015,
+      "loss": 3.786245822906494,
+      "step": 2474
+    },
+    {
+      "epoch": 0.7523368037084884,
+      "grad_norm": 2.35770320892334,
+      "learning_rate": 0.00015,
+      "loss": 3.6131296157836914,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7526407781746334,
+      "grad_norm": 2.6588618755340576,
+      "learning_rate": 0.00015,
+      "loss": 4.021531105041504,
+      "step": 2476
+    },
+    {
+      "epoch": 0.7529447526407782,
+      "grad_norm": 2.941587448120117,
+      "learning_rate": 0.00015,
+      "loss": 4.196323871612549,
+      "step": 2477
+    },
+    {
+      "epoch": 0.753248727106923,
+      "grad_norm": 1.953205943107605,
+      "learning_rate": 0.00015,
+      "loss": 3.6473865509033203,
+      "step": 2478
+    },
+    {
+      "epoch": 0.7535527015730679,
+      "grad_norm": 1.963150978088379,
+      "learning_rate": 0.00015,
+      "loss": 3.782966136932373,
+      "step": 2479
+    },
+    {
+      "epoch": 0.7538566760392127,
+      "grad_norm": 1.7462246417999268,
+      "learning_rate": 0.00015,
+      "loss": 3.7924442291259766,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7541606505053575,
+      "grad_norm": 1.897908091545105,
+      "learning_rate": 0.00015,
+      "loss": 4.140430450439453,
+      "step": 2481
+    },
+    {
+      "epoch": 0.7544646249715024,
+      "grad_norm": 2.0684754848480225,
+      "learning_rate": 0.00015,
+      "loss": 3.8213205337524414,
+      "step": 2482
+    },
+    {
+      "epoch": 0.7547685994376473,
+      "grad_norm": 1.8376339673995972,
+      "learning_rate": 0.00015,
+      "loss": 3.49650239944458,
+      "step": 2483
+    },
+    {
+      "epoch": 0.7550725739037921,
+      "grad_norm": 2.431823968887329,
+      "learning_rate": 0.00015,
+      "loss": 3.965762138366699,
+      "step": 2484
+    },
+    {
+      "epoch": 0.755376548369937,
+      "grad_norm": 2.0553362369537354,
+      "learning_rate": 0.00015,
+      "loss": 3.4433631896972656,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7556805228360818,
+      "grad_norm": 1.7035285234451294,
+      "learning_rate": 0.00015,
+      "loss": 3.7297518253326416,
+      "step": 2486
+    },
+    {
+      "epoch": 0.7559844973022266,
+      "grad_norm": 1.693340539932251,
+      "learning_rate": 0.00015,
+      "loss": 3.8057453632354736,
+      "step": 2487
+    },
+    {
+      "epoch": 0.7562884717683714,
+      "grad_norm": 2.3725388050079346,
+      "learning_rate": 0.00015,
+      "loss": 3.9326882362365723,
+      "step": 2488
+    },
+    {
+      "epoch": 0.7565924462345163,
+      "grad_norm": 2.4714536666870117,
+      "learning_rate": 0.00015,
+      "loss": 3.5451650619506836,
+      "step": 2489
+    },
+    {
+      "epoch": 0.7568964207006611,
+      "grad_norm": 1.9930287599563599,
+      "learning_rate": 0.00015,
+      "loss": 3.8311970233917236,
+      "step": 2490
+    },
+    {
+      "epoch": 0.757200395166806,
+      "grad_norm": 2.000913619995117,
+      "learning_rate": 0.00015,
+      "loss": 3.4476475715637207,
+      "step": 2491
+    },
+    {
+      "epoch": 0.7575043696329509,
+      "grad_norm": 1.7586987018585205,
+      "learning_rate": 0.00015,
+      "loss": 3.6873939037323,
+      "step": 2492
+    },
+    {
+      "epoch": 0.7578083440990957,
+      "grad_norm": 2.041945695877075,
+      "learning_rate": 0.00015,
+      "loss": 3.55676531791687,
+      "step": 2493
+    },
+    {
+      "epoch": 0.7581123185652405,
+      "grad_norm": 2.5155527591705322,
+      "learning_rate": 0.00015,
+      "loss": 4.032420635223389,
+      "step": 2494
+    },
+    {
+      "epoch": 0.7584162930313854,
+      "grad_norm": 1.9654513597488403,
+      "learning_rate": 0.00015,
+      "loss": 3.8009448051452637,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7587202674975302,
+      "grad_norm": 2.916109800338745,
+      "learning_rate": 0.00015,
+      "loss": 3.475721597671509,
+      "step": 2496
+    },
+    {
+      "epoch": 0.759024241963675,
+      "grad_norm": 1.8387186527252197,
+      "learning_rate": 0.00015,
+      "loss": 3.6381826400756836,
+      "step": 2497
+    },
+    {
+      "epoch": 0.7593282164298198,
+      "grad_norm": 1.9014939069747925,
+      "learning_rate": 0.00015,
+      "loss": 3.695781946182251,
+      "step": 2498
+    },
+    {
+      "epoch": 0.7596321908959648,
+      "grad_norm": 1.8259621858596802,
+      "learning_rate": 0.00015,
+      "loss": 4.005601406097412,
+      "step": 2499
+    },
+    {
+      "epoch": 0.7599361653621096,
+      "grad_norm": 1.8048176765441895,
+      "learning_rate": 0.00015,
+      "loss": 3.535665512084961,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7602401398282544,
+      "grad_norm": 2.7526018619537354,
+      "learning_rate": 0.00015,
+      "loss": 3.6382479667663574,
+      "step": 2501
+    },
+    {
+      "epoch": 0.7605441142943993,
+      "grad_norm": 2.5160279273986816,
+      "learning_rate": 0.00015,
+      "loss": 3.8469319343566895,
+      "step": 2502
+    },
+    {
+      "epoch": 0.7608480887605441,
+      "grad_norm": 4.325017929077148,
+      "learning_rate": 0.00015,
+      "loss": 3.315335750579834,
+      "step": 2503
+    },
+    {
+      "epoch": 0.7611520632266889,
+      "grad_norm": 2.0879993438720703,
+      "learning_rate": 0.00015,
+      "loss": 3.7138195037841797,
+      "step": 2504
+    },
+    {
+      "epoch": 0.7614560376928338,
+      "grad_norm": 1.9412068128585815,
+      "learning_rate": 0.00015,
+      "loss": 3.7700281143188477,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7617600121589786,
+      "grad_norm": 1.4645812511444092,
+      "learning_rate": 0.00015,
+      "loss": 3.711348056793213,
+      "step": 2506
+    },
+    {
+      "epoch": 0.7620639866251235,
+      "grad_norm": 2.927170991897583,
+      "learning_rate": 0.00015,
+      "loss": 3.8271689414978027,
+      "step": 2507
+    },
+    {
+      "epoch": 0.7623679610912684,
+      "grad_norm": 1.8109536170959473,
+      "learning_rate": 0.00015,
+      "loss": 3.904282569885254,
+      "step": 2508
+    },
+    {
+      "epoch": 0.7626719355574132,
+      "grad_norm": 1.913558840751648,
+      "learning_rate": 0.00015,
+      "loss": 3.809675693511963,
+      "step": 2509
+    },
+    {
+      "epoch": 0.762975910023558,
+      "grad_norm": 1.8769218921661377,
+      "learning_rate": 0.00015,
+      "loss": 3.8619384765625,
+      "step": 2510
+    },
+    {
+      "epoch": 0.7632798844897029,
+      "grad_norm": 1.8363088369369507,
+      "learning_rate": 0.00015,
+      "loss": 4.1027326583862305,
+      "step": 2511
+    },
+    {
+      "epoch": 0.7635838589558477,
+      "grad_norm": 2.1616458892822266,
+      "learning_rate": 0.00015,
+      "loss": 3.6998238563537598,
+      "step": 2512
+    },
+    {
+      "epoch": 0.7638878334219925,
+      "grad_norm": 1.8330612182617188,
+      "learning_rate": 0.00015,
+      "loss": 3.7776288986206055,
+      "step": 2513
+    },
+    {
+      "epoch": 0.7641918078881373,
+      "grad_norm": 2.1166679859161377,
+      "learning_rate": 0.00015,
+      "loss": 4.258569717407227,
+      "step": 2514
+    },
+    {
+      "epoch": 0.7644957823542823,
+      "grad_norm": 2.2348461151123047,
+      "learning_rate": 0.00015,
+      "loss": 3.829911708831787,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7647997568204271,
+      "grad_norm": 2.3700499534606934,
+      "learning_rate": 0.00015,
+      "loss": 3.928175926208496,
+      "step": 2516
+    },
+    {
+      "epoch": 0.7651037312865719,
+      "grad_norm": 2.0304105281829834,
+      "learning_rate": 0.00015,
+      "loss": 4.095639228820801,
+      "step": 2517
+    },
+    {
+      "epoch": 0.7654077057527168,
+      "grad_norm": 2.03615403175354,
+      "learning_rate": 0.00015,
+      "loss": 3.799692392349243,
+      "step": 2518
+    },
+    {
+      "epoch": 0.7657116802188616,
+      "grad_norm": 2.5357017517089844,
+      "learning_rate": 0.00015,
+      "loss": 3.674919605255127,
+      "step": 2519
+    },
+    {
+      "epoch": 0.7660156546850064,
+      "grad_norm": 1.7603638172149658,
+      "learning_rate": 0.00015,
+      "loss": 3.7336983680725098,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7663196291511513,
+      "grad_norm": 1.566463589668274,
+      "learning_rate": 0.00015,
+      "loss": 4.027907371520996,
+      "step": 2521
+    },
+    {
+      "epoch": 0.7666236036172962,
+      "grad_norm": 2.1519267559051514,
+      "learning_rate": 0.00015,
+      "loss": 3.522902011871338,
+      "step": 2522
+    },
+    {
+      "epoch": 0.766927578083441,
+      "grad_norm": 2.1586437225341797,
+      "learning_rate": 0.00015,
+      "loss": 3.847123146057129,
+      "step": 2523
+    },
+    {
+      "epoch": 0.7672315525495859,
+      "grad_norm": 1.7018686532974243,
+      "learning_rate": 0.00015,
+      "loss": 3.781999111175537,
+      "step": 2524
+    },
+    {
+      "epoch": 0.7675355270157307,
+      "grad_norm": 2.105112314224243,
+      "learning_rate": 0.00015,
+      "loss": 3.750377655029297,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7678395014818755,
+      "grad_norm": 2.5717575550079346,
+      "learning_rate": 0.00015,
+      "loss": 3.5032193660736084,
+      "step": 2526
+    },
+    {
+      "epoch": 0.7681434759480203,
+      "grad_norm": 1.7363030910491943,
+      "learning_rate": 0.00015,
+      "loss": 3.5548040866851807,
+      "step": 2527
+    },
+    {
+      "epoch": 0.7684474504141652,
+      "grad_norm": 2.2118418216705322,
+      "learning_rate": 0.00015,
+      "loss": 4.290332317352295,
+      "step": 2528
+    },
+    {
+      "epoch": 0.76875142488031,
+      "grad_norm": 2.0474693775177,
+      "learning_rate": 0.00015,
+      "loss": 3.4323220252990723,
+      "step": 2529
+    },
+    {
+      "epoch": 0.769055399346455,
+      "grad_norm": 1.8353263139724731,
+      "learning_rate": 0.00015,
+      "loss": 3.855989456176758,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7693593738125998,
+      "grad_norm": 1.9186216592788696,
+      "learning_rate": 0.00015,
+      "loss": 3.828808069229126,
+      "step": 2531
+    },
+    {
+      "epoch": 0.7696633482787446,
+      "grad_norm": 1.7198150157928467,
+      "learning_rate": 0.00015,
+      "loss": 3.6927309036254883,
+      "step": 2532
+    },
+    {
+      "epoch": 0.7699673227448894,
+      "grad_norm": 2.261291742324829,
+      "learning_rate": 0.00015,
+      "loss": 3.633824348449707,
+      "step": 2533
+    },
+    {
+      "epoch": 0.7702712972110343,
+      "grad_norm": 2.4790422916412354,
+      "learning_rate": 0.00015,
+      "loss": 3.881030797958374,
+      "step": 2534
+    },
+    {
+      "epoch": 0.7705752716771791,
+      "grad_norm": 2.2233150005340576,
+      "learning_rate": 0.00015,
+      "loss": 3.8003482818603516,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7708792461433239,
+      "grad_norm": 2.480051040649414,
+      "learning_rate": 0.00015,
+      "loss": 3.646253824234009,
+      "step": 2536
+    },
+    {
+      "epoch": 0.7711832206094688,
+      "grad_norm": 2.068129062652588,
+      "learning_rate": 0.00015,
+      "loss": 3.857926368713379,
+      "step": 2537
+    },
+    {
+      "epoch": 0.7714871950756137,
+      "grad_norm": 2.3867602348327637,
+      "learning_rate": 0.00015,
+      "loss": 3.840299367904663,
+      "step": 2538
+    },
+    {
+      "epoch": 0.7717911695417585,
+      "grad_norm": 1.8332921266555786,
+      "learning_rate": 0.00015,
+      "loss": 3.8716061115264893,
+      "step": 2539
+    },
+    {
+      "epoch": 0.7720951440079034,
+      "grad_norm": 1.984262228012085,
+      "learning_rate": 0.00015,
+      "loss": 3.967808485031128,
+      "step": 2540
+    },
+    {
+      "epoch": 0.7723991184740482,
+      "grad_norm": 1.6272228956222534,
+      "learning_rate": 0.00015,
+      "loss": 3.795588731765747,
+      "step": 2541
+    },
+    {
+      "epoch": 0.772703092940193,
+      "grad_norm": 1.6950212717056274,
+      "learning_rate": 0.00015,
+      "loss": 3.694403648376465,
+      "step": 2542
+    },
+    {
+      "epoch": 0.7730070674063378,
+      "grad_norm": 2.1847684383392334,
+      "learning_rate": 0.00015,
+      "loss": 3.9573276042938232,
+      "step": 2543
+    },
+    {
+      "epoch": 0.7733110418724827,
+      "grad_norm": 1.620031714439392,
+      "learning_rate": 0.00015,
+      "loss": 3.8277621269226074,
+      "step": 2544
+    },
+    {
+      "epoch": 0.7736150163386275,
+      "grad_norm": 1.8515030145645142,
+      "learning_rate": 0.00015,
+      "loss": 3.6059608459472656,
+      "step": 2545
+    },
+    {
+      "epoch": 0.7739189908047724,
+      "grad_norm": 2.160086154937744,
+      "learning_rate": 0.00015,
+      "loss": 3.875939130783081,
+      "step": 2546
+    },
+    {
+      "epoch": 0.7742229652709173,
+      "grad_norm": 2.016993761062622,
+      "learning_rate": 0.00015,
+      "loss": 3.8807120323181152,
+      "step": 2547
+    },
+    {
+      "epoch": 0.7745269397370621,
+      "grad_norm": 1.6484489440917969,
+      "learning_rate": 0.00015,
+      "loss": 3.7898430824279785,
+      "step": 2548
+    },
+    {
+      "epoch": 0.7748309142032069,
+      "grad_norm": 1.6960625648498535,
+      "learning_rate": 0.00015,
+      "loss": 3.5670583248138428,
+      "step": 2549
+    },
+    {
+      "epoch": 0.7751348886693518,
+      "grad_norm": 1.8637651205062866,
+      "learning_rate": 0.00015,
+      "loss": 3.744180679321289,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7754388631354966,
+      "grad_norm": 1.7673426866531372,
+      "learning_rate": 0.00015,
+      "loss": 3.859964370727539,
+      "step": 2551
+    },
+    {
+      "epoch": 0.7757428376016414,
+      "grad_norm": 1.7987306118011475,
+      "learning_rate": 0.00015,
+      "loss": 3.856010913848877,
+      "step": 2552
+    },
+    {
+      "epoch": 0.7760468120677863,
+      "grad_norm": 2.0558290481567383,
+      "learning_rate": 0.00015,
+      "loss": 3.7617063522338867,
+      "step": 2553
+    },
+    {
+      "epoch": 0.7763507865339312,
+      "grad_norm": 1.9272178411483765,
+      "learning_rate": 0.00015,
+      "loss": 3.5718746185302734,
+      "step": 2554
+    },
+    {
+      "epoch": 0.776654761000076,
+      "grad_norm": 1.9480279684066772,
+      "learning_rate": 0.00015,
+      "loss": 3.9146156311035156,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7769587354662209,
+      "grad_norm": 2.1273930072784424,
+      "learning_rate": 0.00015,
+      "loss": 3.769538640975952,
+      "step": 2556
+    },
+    {
+      "epoch": 0.7772627099323657,
+      "grad_norm": 1.8962479829788208,
+      "learning_rate": 0.00015,
+      "loss": 3.752241611480713,
+      "step": 2557
+    },
+    {
+      "epoch": 0.7775666843985105,
+      "grad_norm": 1.8411680459976196,
+      "learning_rate": 0.00015,
+      "loss": 3.821099042892456,
+      "step": 2558
+    },
+    {
+      "epoch": 0.7778706588646553,
+      "grad_norm": 1.8188793659210205,
+      "learning_rate": 0.00015,
+      "loss": 3.8110456466674805,
+      "step": 2559
+    },
+    {
+      "epoch": 0.7781746333308002,
+      "grad_norm": 1.987066626548767,
+      "learning_rate": 0.00015,
+      "loss": 3.62813663482666,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7784786077969451,
+      "grad_norm": 1.8750032186508179,
+      "learning_rate": 0.00015,
+      "loss": 3.8542981147766113,
+      "step": 2561
+    },
+    {
+      "epoch": 0.7787825822630899,
+      "grad_norm": 2.578000068664551,
+      "learning_rate": 0.00015,
+      "loss": 3.7646806240081787,
+      "step": 2562
+    },
+    {
+      "epoch": 0.7790865567292348,
+      "grad_norm": 2.1084201335906982,
+      "learning_rate": 0.00015,
+      "loss": 3.7189629077911377,
+      "step": 2563
+    },
+    {
+      "epoch": 0.7793905311953796,
+      "grad_norm": 2.1420412063598633,
+      "learning_rate": 0.00015,
+      "loss": 3.517512798309326,
+      "step": 2564
+    },
+    {
+      "epoch": 0.7796945056615244,
+      "grad_norm": 2.2180676460266113,
+      "learning_rate": 0.00015,
+      "loss": 4.179899215698242,
+      "step": 2565
+    },
+    {
+      "epoch": 0.7799984801276693,
+      "grad_norm": 2.4641225337982178,
+      "learning_rate": 0.00015,
+      "loss": 3.930799961090088,
+      "step": 2566
+    },
+    {
+      "epoch": 0.7803024545938141,
+      "grad_norm": 1.9050483703613281,
+      "learning_rate": 0.00015,
+      "loss": 3.671574592590332,
+      "step": 2567
+    },
+    {
+      "epoch": 0.7806064290599589,
+      "grad_norm": 1.9253487586975098,
+      "learning_rate": 0.00015,
+      "loss": 3.8670806884765625,
+      "step": 2568
+    },
+    {
+      "epoch": 0.7809104035261039,
+      "grad_norm": 2.0895090103149414,
+      "learning_rate": 0.00015,
+      "loss": 3.6279404163360596,
+      "step": 2569
+    },
+    {
+      "epoch": 0.7812143779922487,
+      "grad_norm": 1.9596447944641113,
+      "learning_rate": 0.00015,
+      "loss": 3.9193365573883057,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7815183524583935,
+      "grad_norm": 2.094878911972046,
+      "learning_rate": 0.00015,
+      "loss": 3.8763070106506348,
+      "step": 2571
+    },
+    {
+      "epoch": 0.7818223269245383,
+      "grad_norm": 1.6455186605453491,
+      "learning_rate": 0.00015,
+      "loss": 3.91217303276062,
+      "step": 2572
+    },
+    {
+      "epoch": 0.7821263013906832,
+      "grad_norm": 1.804026484489441,
+      "learning_rate": 0.00015,
+      "loss": 3.6970465183258057,
+      "step": 2573
+    },
+    {
+      "epoch": 0.782430275856828,
+      "grad_norm": 2.0372157096862793,
+      "learning_rate": 0.00015,
+      "loss": 3.5757205486297607,
+      "step": 2574
+    },
+    {
+      "epoch": 0.7827342503229728,
+      "grad_norm": 2.438642740249634,
+      "learning_rate": 0.00015,
+      "loss": 3.645099401473999,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7830382247891177,
+      "grad_norm": 2.557295083999634,
+      "learning_rate": 0.00015,
+      "loss": 3.882017135620117,
+      "step": 2576
+    },
+    {
+      "epoch": 0.7833421992552626,
+      "grad_norm": 1.7348712682724,
+      "learning_rate": 0.00015,
+      "loss": 4.049770355224609,
+      "step": 2577
+    },
+    {
+      "epoch": 0.7836461737214074,
+      "grad_norm": 2.4487221240997314,
+      "learning_rate": 0.00015,
+      "loss": 3.9012250900268555,
+      "step": 2578
+    },
+    {
+      "epoch": 0.7839501481875523,
+      "grad_norm": 1.7075611352920532,
+      "learning_rate": 0.00015,
+      "loss": 3.4436540603637695,
+      "step": 2579
+    },
+    {
+      "epoch": 0.7842541226536971,
+      "grad_norm": 2.1210622787475586,
+      "learning_rate": 0.00015,
+      "loss": 3.8098888397216797,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7845580971198419,
+      "grad_norm": 1.8724398612976074,
+      "learning_rate": 0.00015,
+      "loss": 3.6263089179992676,
+      "step": 2581
+    },
+    {
+      "epoch": 0.7848620715859868,
+      "grad_norm": 1.9030404090881348,
+      "learning_rate": 0.00015,
+      "loss": 3.7149758338928223,
+      "step": 2582
+    },
+    {
+      "epoch": 0.7851660460521316,
+      "grad_norm": 1.9976400136947632,
+      "learning_rate": 0.00015,
+      "loss": 3.855456590652466,
+      "step": 2583
+    },
+    {
+      "epoch": 0.7854700205182764,
+      "grad_norm": 2.2151505947113037,
+      "learning_rate": 0.00015,
+      "loss": 3.439734697341919,
+      "step": 2584
+    },
+    {
+      "epoch": 0.7857739949844214,
+      "grad_norm": 1.9165127277374268,
+      "learning_rate": 0.00015,
+      "loss": 3.7489566802978516,
+      "step": 2585
+    },
+    {
+      "epoch": 0.7860779694505662,
+      "grad_norm": 1.9534214735031128,
+      "learning_rate": 0.00015,
+      "loss": 3.718020439147949,
+      "step": 2586
+    },
+    {
+      "epoch": 0.786381943916711,
+      "grad_norm": 2.248286724090576,
+      "learning_rate": 0.00015,
+      "loss": 3.7337379455566406,
+      "step": 2587
+    },
+    {
+      "epoch": 0.7866859183828558,
+      "grad_norm": 2.4801974296569824,
+      "learning_rate": 0.00015,
+      "loss": 3.8031411170959473,
+      "step": 2588
+    },
+    {
+      "epoch": 0.7869898928490007,
+      "grad_norm": 2.380828380584717,
+      "learning_rate": 0.00015,
+      "loss": 3.8999812602996826,
+      "step": 2589
+    },
+    {
+      "epoch": 0.7872938673151455,
+      "grad_norm": 1.9307570457458496,
+      "learning_rate": 0.00015,
+      "loss": 3.6343131065368652,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7875978417812903,
+      "grad_norm": 2.4790520668029785,
+      "learning_rate": 0.00015,
+      "loss": 4.111806869506836,
+      "step": 2591
+    },
+    {
+      "epoch": 0.7879018162474352,
+      "grad_norm": 2.140328884124756,
+      "learning_rate": 0.00015,
+      "loss": 3.5890822410583496,
+      "step": 2592
+    },
+    {
+      "epoch": 0.7882057907135801,
+      "grad_norm": 2.2938222885131836,
+      "learning_rate": 0.00015,
+      "loss": 4.138959884643555,
+      "step": 2593
+    },
+    {
+      "epoch": 0.7885097651797249,
+      "grad_norm": 1.849379539489746,
+      "learning_rate": 0.00015,
+      "loss": 3.7599496841430664,
+      "step": 2594
+    },
+    {
+      "epoch": 0.7888137396458698,
+      "grad_norm": 2.560744047164917,
+      "learning_rate": 0.00015,
+      "loss": 3.4714314937591553,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7891177141120146,
+      "grad_norm": 2.4190480709075928,
+      "learning_rate": 0.00015,
+      "loss": 3.4989511966705322,
+      "step": 2596
+    },
+    {
+      "epoch": 0.7894216885781594,
+      "grad_norm": 2.064272403717041,
+      "learning_rate": 0.00015,
+      "loss": 3.692965030670166,
+      "step": 2597
+    },
+    {
+      "epoch": 0.7897256630443042,
+      "grad_norm": 2.1233930587768555,
+      "learning_rate": 0.00015,
+      "loss": 3.829005241394043,
+      "step": 2598
+    },
+    {
+      "epoch": 0.7900296375104491,
+      "grad_norm": 2.949523687362671,
+      "learning_rate": 0.00015,
+      "loss": 3.8187804222106934,
+      "step": 2599
+    },
+    {
+      "epoch": 0.790333611976594,
+      "grad_norm": 2.121873140335083,
+      "learning_rate": 0.00015,
+      "loss": 3.785015344619751,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7906375864427388,
+      "grad_norm": 1.943771243095398,
+      "learning_rate": 0.00015,
+      "loss": 3.495471477508545,
+      "step": 2601
+    },
+    {
+      "epoch": 0.7909415609088837,
+      "grad_norm": 1.9529200792312622,
+      "learning_rate": 0.00015,
+      "loss": 3.4498891830444336,
+      "step": 2602
+    },
+    {
+      "epoch": 0.7912455353750285,
+      "grad_norm": 1.780472755432129,
+      "learning_rate": 0.00015,
+      "loss": 3.8255631923675537,
+      "step": 2603
+    },
+    {
+      "epoch": 0.7915495098411733,
+      "grad_norm": 2.216336250305176,
+      "learning_rate": 0.00015,
+      "loss": 3.540290594100952,
+      "step": 2604
+    },
+    {
+      "epoch": 0.7918534843073182,
+      "grad_norm": 2.4086997509002686,
+      "learning_rate": 0.00015,
+      "loss": 3.906144380569458,
+      "step": 2605
+    },
+    {
+      "epoch": 0.792157458773463,
+      "grad_norm": 2.0140678882598877,
+      "learning_rate": 0.00015,
+      "loss": 4.003551006317139,
+      "step": 2606
+    },
+    {
+      "epoch": 0.7924614332396078,
+      "grad_norm": 1.994866967201233,
+      "learning_rate": 0.00015,
+      "loss": 3.902479648590088,
+      "step": 2607
+    },
+    {
+      "epoch": 0.7927654077057528,
+      "grad_norm": 2.159229278564453,
+      "learning_rate": 0.00015,
+      "loss": 3.5481858253479004,
+      "step": 2608
+    },
+    {
+      "epoch": 0.7930693821718976,
+      "grad_norm": 2.9118597507476807,
+      "learning_rate": 0.00015,
+      "loss": 4.060391426086426,
+      "step": 2609
+    },
+    {
+      "epoch": 0.7933733566380424,
+      "grad_norm": 2.32466983795166,
+      "learning_rate": 0.00015,
+      "loss": 3.5907139778137207,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7936773311041873,
+      "grad_norm": 2.1032469272613525,
+      "learning_rate": 0.00015,
+      "loss": 3.5641980171203613,
+      "step": 2611
+    },
+    {
+      "epoch": 0.7939813055703321,
+      "grad_norm": 2.1163992881774902,
+      "learning_rate": 0.00015,
+      "loss": 3.469921588897705,
+      "step": 2612
+    },
+    {
+      "epoch": 0.7942852800364769,
+      "grad_norm": 1.8649039268493652,
+      "learning_rate": 0.00015,
+      "loss": 3.2529518604278564,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7945892545026217,
+      "grad_norm": 1.8021823167800903,
+      "learning_rate": 0.00015,
+      "loss": 3.690062999725342,
+      "step": 2614
+    },
+    {
+      "epoch": 0.7948932289687666,
+      "grad_norm": 2.664412498474121,
+      "learning_rate": 0.00015,
+      "loss": 3.775620698928833,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7951972034349115,
+      "grad_norm": 3.3621675968170166,
+      "learning_rate": 0.00015,
+      "loss": 3.8509316444396973,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7955011779010563,
+      "grad_norm": 1.7779884338378906,
+      "learning_rate": 0.00015,
+      "loss": 3.50534987449646,
+      "step": 2617
+    },
+    {
+      "epoch": 0.7958051523672012,
+      "grad_norm": 1.8307665586471558,
+      "learning_rate": 0.00015,
+      "loss": 3.6658554077148438,
+      "step": 2618
+    },
+    {
+      "epoch": 0.796109126833346,
+      "grad_norm": 2.4320406913757324,
+      "learning_rate": 0.00015,
+      "loss": 4.053743362426758,
+      "step": 2619
+    },
+    {
+      "epoch": 0.7964131012994908,
+      "grad_norm": 1.9290564060211182,
+      "learning_rate": 0.00015,
+      "loss": 3.6696338653564453,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7967170757656357,
+      "grad_norm": 2.003065347671509,
+      "learning_rate": 0.00015,
+      "loss": 3.7810425758361816,
+      "step": 2621
+    },
+    {
+      "epoch": 0.7970210502317805,
+      "grad_norm": 2.7063088417053223,
+      "learning_rate": 0.00015,
+      "loss": 3.7923765182495117,
+      "step": 2622
+    },
+    {
+      "epoch": 0.7973250246979253,
+      "grad_norm": 2.245702028274536,
+      "learning_rate": 0.00015,
+      "loss": 3.577545642852783,
+      "step": 2623
+    },
+    {
+      "epoch": 0.7976289991640703,
+      "grad_norm": 1.8542309999465942,
+      "learning_rate": 0.00015,
+      "loss": 3.730398654937744,
+      "step": 2624
+    },
+    {
+      "epoch": 0.7979329736302151,
+      "grad_norm": 2.2563438415527344,
+      "learning_rate": 0.00015,
+      "loss": 3.6075868606567383,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7982369480963599,
+      "grad_norm": 1.5706051588058472,
+      "learning_rate": 0.00015,
+      "loss": 3.5569097995758057,
+      "step": 2626
+    },
+    {
+      "epoch": 0.7985409225625048,
+      "grad_norm": 2.0690758228302,
+      "learning_rate": 0.00015,
+      "loss": 3.8182945251464844,
+      "step": 2627
+    },
+    {
+      "epoch": 0.7988448970286496,
+      "grad_norm": 2.3047566413879395,
+      "learning_rate": 0.00015,
+      "loss": 3.4351682662963867,
+      "step": 2628
+    },
+    {
+      "epoch": 0.7991488714947944,
+      "grad_norm": 2.03216290473938,
+      "learning_rate": 0.00015,
+      "loss": 3.638763427734375,
+      "step": 2629
+    },
+    {
+      "epoch": 0.7994528459609392,
+      "grad_norm": 2.8470420837402344,
+      "learning_rate": 0.00015,
+      "loss": 3.8779053688049316,
+      "step": 2630
+    },
+    {
+      "epoch": 0.7997568204270841,
+      "grad_norm": 2.057551860809326,
+      "learning_rate": 0.00015,
+      "loss": 3.627102851867676,
+      "step": 2631
+    },
+    {
+      "epoch": 0.800060794893229,
+      "grad_norm": 2.17386531829834,
+      "learning_rate": 0.00015,
+      "loss": 3.782426595687866,
+      "step": 2632
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3290,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 329,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.243410134426452e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}