{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9266409266409267,
  "eval_steps": 500,
  "global_step": 1800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005148005148005148,
      "grad_norm": 19.51059658759793,
      "learning_rate": 7.718696397941682e-07,
      "loss": 8.2552,
      "step": 10
    },
    {
      "epoch": 0.010296010296010296,
      "grad_norm": 19.70146681963939,
      "learning_rate": 1.6295025728987994e-06,
      "loss": 8.168,
      "step": 20
    },
    {
      "epoch": 0.015444015444015444,
      "grad_norm": 9.795523599582376,
      "learning_rate": 2.4871355060034306e-06,
      "loss": 7.5953,
      "step": 30
    },
    {
      "epoch": 0.02059202059202059,
      "grad_norm": 5.9480235723668144,
      "learning_rate": 3.344768439108062e-06,
      "loss": 6.8507,
      "step": 40
    },
    {
      "epoch": 0.02574002574002574,
      "grad_norm": 3.5099463285447223,
      "learning_rate": 4.202401372212693e-06,
      "loss": 6.3398,
      "step": 50
    },
    {
      "epoch": 0.03088803088803089,
      "grad_norm": 2.90890000793416,
      "learning_rate": 5.060034305317325e-06,
      "loss": 5.9988,
      "step": 60
    },
    {
      "epoch": 0.036036036036036036,
      "grad_norm": 3.3978017812772445,
      "learning_rate": 5.9176672384219555e-06,
      "loss": 5.6503,
      "step": 70
    },
    {
      "epoch": 0.04118404118404118,
      "grad_norm": 4.103088821630264,
      "learning_rate": 6.775300171526587e-06,
      "loss": 5.1289,
      "step": 80
    },
    {
      "epoch": 0.04633204633204633,
      "grad_norm": 4.487842621805789,
      "learning_rate": 7.632933104631218e-06,
      "loss": 4.511,
      "step": 90
    },
    {
      "epoch": 0.05148005148005148,
      "grad_norm": 7.491973587682773,
      "learning_rate": 8.49056603773585e-06,
      "loss": 3.9137,
      "step": 100
    },
    {
      "epoch": 0.05662805662805663,
      "grad_norm": 7.323430680232261,
      "learning_rate": 9.348198970840481e-06,
      "loss": 3.4784,
      "step": 110
    },
    {
      "epoch": 0.06177606177606178,
      "grad_norm": 6.619474879048035,
      "learning_rate": 1.0205831903945113e-05,
      "loss": 3.2067,
      "step": 120
    },
    {
      "epoch": 0.06692406692406692,
      "grad_norm": 3.867444417417378,
      "learning_rate": 1.1063464837049744e-05,
      "loss": 3.0386,
      "step": 130
    },
    {
      "epoch": 0.07207207207207207,
      "grad_norm": 4.117034943974073,
      "learning_rate": 1.1921097770154374e-05,
      "loss": 2.9159,
      "step": 140
    },
    {
      "epoch": 0.07722007722007722,
      "grad_norm": 6.009222032591652,
      "learning_rate": 1.2778730703259007e-05,
      "loss": 2.796,
      "step": 150
    },
    {
      "epoch": 0.08236808236808237,
      "grad_norm": 11.57505334622991,
      "learning_rate": 1.3636363636363637e-05,
      "loss": 2.7077,
      "step": 160
    },
    {
      "epoch": 0.08751608751608751,
      "grad_norm": 11.428627421387615,
      "learning_rate": 1.4493996569468268e-05,
      "loss": 2.6403,
      "step": 170
    },
    {
      "epoch": 0.09266409266409266,
      "grad_norm": 10.905546843716907,
      "learning_rate": 1.53516295025729e-05,
      "loss": 2.5735,
      "step": 180
    },
    {
      "epoch": 0.0978120978120978,
      "grad_norm": 8.661551806496005,
      "learning_rate": 1.620926243567753e-05,
      "loss": 2.5078,
      "step": 190
    },
    {
      "epoch": 0.10296010296010295,
      "grad_norm": 9.940151278870877,
      "learning_rate": 1.7066895368782164e-05,
      "loss": 2.4567,
      "step": 200
    },
    {
      "epoch": 0.10810810810810811,
      "grad_norm": 13.091142904495205,
      "learning_rate": 1.7924528301886792e-05,
      "loss": 2.3953,
      "step": 210
    },
    {
      "epoch": 0.11325611325611326,
      "grad_norm": 18.06611865183751,
      "learning_rate": 1.8782161234991423e-05,
      "loss": 2.329,
      "step": 220
    },
    {
      "epoch": 0.11840411840411841,
      "grad_norm": 16.145152244084983,
      "learning_rate": 1.9639794168096058e-05,
      "loss": 2.2691,
      "step": 230
    },
    {
      "epoch": 0.12355212355212356,
      "grad_norm": 35.1534299592601,
      "learning_rate": 2.0497427101200686e-05,
      "loss": 2.2524,
      "step": 240
    },
    {
      "epoch": 0.1287001287001287,
      "grad_norm": 14.465207787142148,
      "learning_rate": 2.1355060034305318e-05,
      "loss": 2.1974,
      "step": 250
    },
    {
      "epoch": 0.13384813384813385,
      "grad_norm": 11.353218963690644,
      "learning_rate": 2.2212692967409952e-05,
      "loss": 2.0873,
      "step": 260
    },
    {
      "epoch": 0.138996138996139,
      "grad_norm": 12.277610068118626,
      "learning_rate": 2.307032590051458e-05,
      "loss": 2.0068,
      "step": 270
    },
    {
      "epoch": 0.14414414414414414,
      "grad_norm": 10.505580540439455,
      "learning_rate": 2.3927958833619212e-05,
      "loss": 1.8888,
      "step": 280
    },
    {
      "epoch": 0.1492921492921493,
      "grad_norm": 14.960676861450345,
      "learning_rate": 2.4785591766723843e-05,
      "loss": 1.7894,
      "step": 290
    },
    {
      "epoch": 0.15444015444015444,
      "grad_norm": 8.613188849395,
      "learning_rate": 2.564322469982847e-05,
      "loss": 1.6967,
      "step": 300
    },
    {
      "epoch": 0.15958815958815958,
      "grad_norm": 6.959841307105666,
      "learning_rate": 2.6500857632933106e-05,
      "loss": 1.6019,
      "step": 310
    },
    {
      "epoch": 0.16473616473616473,
      "grad_norm": 5.864669793953841,
      "learning_rate": 2.7358490566037738e-05,
      "loss": 1.5326,
      "step": 320
    },
    {
      "epoch": 0.16988416988416988,
      "grad_norm": 8.165848406130657,
      "learning_rate": 2.8216123499142366e-05,
      "loss": 1.4842,
      "step": 330
    },
    {
      "epoch": 0.17503217503217502,
      "grad_norm": 6.747629987576382,
      "learning_rate": 2.9073756432247e-05,
      "loss": 1.426,
      "step": 340
    },
    {
      "epoch": 0.18018018018018017,
      "grad_norm": 4.897733377675901,
      "learning_rate": 2.9931389365351632e-05,
      "loss": 1.3614,
      "step": 350
    },
    {
      "epoch": 0.18532818532818532,
      "grad_norm": 5.330679969567318,
      "learning_rate": 3.078902229845626e-05,
      "loss": 1.3182,
      "step": 360
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 4.722018507382775,
      "learning_rate": 3.16466552315609e-05,
      "loss": 1.26,
      "step": 370
    },
    {
      "epoch": 0.1956241956241956,
      "grad_norm": 6.77470473058928,
      "learning_rate": 3.250428816466552e-05,
      "loss": 1.2128,
      "step": 380
    },
    {
      "epoch": 0.20077220077220076,
      "grad_norm": 4.902136742839717,
      "learning_rate": 3.3361921097770154e-05,
      "loss": 1.1829,
      "step": 390
    },
    {
      "epoch": 0.2059202059202059,
      "grad_norm": 5.515696505859524,
      "learning_rate": 3.421955403087479e-05,
      "loss": 1.1379,
      "step": 400
    },
    {
      "epoch": 0.21106821106821108,
      "grad_norm": 3.9113735830318603,
      "learning_rate": 3.507718696397942e-05,
      "loss": 1.1002,
      "step": 410
    },
    {
      "epoch": 0.21621621621621623,
      "grad_norm": 5.8061510583054545,
      "learning_rate": 3.593481989708405e-05,
      "loss": 1.0744,
      "step": 420
    },
    {
      "epoch": 0.22136422136422138,
      "grad_norm": 5.927584495277828,
      "learning_rate": 3.679245283018868e-05,
      "loss": 1.0304,
      "step": 430
    },
    {
      "epoch": 0.22651222651222652,
      "grad_norm": 5.920319878093127,
      "learning_rate": 3.765008576329331e-05,
      "loss": 1.0023,
      "step": 440
    },
    {
      "epoch": 0.23166023166023167,
      "grad_norm": 5.302890318730487,
      "learning_rate": 3.850771869639794e-05,
      "loss": 0.9776,
      "step": 450
    },
    {
      "epoch": 0.23680823680823682,
      "grad_norm": 5.653239199646973,
      "learning_rate": 3.9365351629502574e-05,
      "loss": 0.9482,
      "step": 460
    },
    {
      "epoch": 0.24195624195624196,
      "grad_norm": 4.852345392975872,
      "learning_rate": 4.0222984562607206e-05,
      "loss": 0.9155,
      "step": 470
    },
    {
      "epoch": 0.2471042471042471,
      "grad_norm": 4.499279652551338,
      "learning_rate": 4.108061749571184e-05,
      "loss": 0.8867,
      "step": 480
    },
    {
      "epoch": 0.25225225225225223,
      "grad_norm": 5.181168954723113,
      "learning_rate": 4.193825042881647e-05,
      "loss": 0.8756,
      "step": 490
    },
    {
      "epoch": 0.2574002574002574,
      "grad_norm": 6.255177791414999,
      "learning_rate": 4.27958833619211e-05,
      "loss": 0.8506,
      "step": 500
    },
    {
      "epoch": 0.2625482625482625,
      "grad_norm": 5.396447906024421,
      "learning_rate": 4.365351629502573e-05,
      "loss": 0.8203,
      "step": 510
    },
    {
      "epoch": 0.2676962676962677,
      "grad_norm": 7.371015885385536,
      "learning_rate": 4.451114922813036e-05,
      "loss": 0.8028,
      "step": 520
    },
    {
      "epoch": 0.2728442728442728,
      "grad_norm": 6.198537777815005,
      "learning_rate": 4.5368782161234994e-05,
      "loss": 0.7808,
      "step": 530
    },
    {
      "epoch": 0.277992277992278,
      "grad_norm": 4.485699286277638,
      "learning_rate": 4.6226415094339625e-05,
      "loss": 0.7639,
      "step": 540
    },
    {
      "epoch": 0.28314028314028317,
      "grad_norm": 6.865323396700289,
      "learning_rate": 4.708404802744426e-05,
      "loss": 0.745,
      "step": 550
    },
    {
      "epoch": 0.2882882882882883,
      "grad_norm": 4.718888462403314,
      "learning_rate": 4.794168096054888e-05,
      "loss": 0.7209,
      "step": 560
    },
    {
      "epoch": 0.29343629343629346,
      "grad_norm": 4.887504164875748,
      "learning_rate": 4.879931389365352e-05,
      "loss": 0.7015,
      "step": 570
    },
    {
      "epoch": 0.2985842985842986,
      "grad_norm": 5.372842102076737,
      "learning_rate": 4.965694682675815e-05,
      "loss": 0.6767,
      "step": 580
    },
    {
      "epoch": 0.30373230373230375,
      "grad_norm": 6.766330192939623,
      "learning_rate": 5.0514579759862776e-05,
      "loss": 0.6624,
      "step": 590
    },
    {
      "epoch": 0.3088803088803089,
      "grad_norm": 4.614837742067113,
      "learning_rate": 5.1372212692967414e-05,
      "loss": 0.6405,
      "step": 600
    },
    {
      "epoch": 0.31402831402831405,
      "grad_norm": 4.201337449433587,
      "learning_rate": 5.222984562607204e-05,
      "loss": 0.6194,
      "step": 610
    },
    {
      "epoch": 0.31917631917631917,
      "grad_norm": 5.51280432030841,
      "learning_rate": 5.308747855917667e-05,
      "loss": 0.5951,
      "step": 620
    },
    {
      "epoch": 0.32432432432432434,
      "grad_norm": 5.129963269925849,
      "learning_rate": 5.394511149228131e-05,
      "loss": 0.5777,
      "step": 630
    },
    {
      "epoch": 0.32947232947232946,
      "grad_norm": 4.412209442955385,
      "learning_rate": 5.480274442538593e-05,
      "loss": 0.5642,
      "step": 640
    },
    {
      "epoch": 0.33462033462033464,
      "grad_norm": 4.336638446576286,
      "learning_rate": 5.5660377358490564e-05,
      "loss": 0.5426,
      "step": 650
    },
    {
      "epoch": 0.33976833976833976,
      "grad_norm": 3.556283909809826,
      "learning_rate": 5.65180102915952e-05,
      "loss": 0.5112,
      "step": 660
    },
    {
      "epoch": 0.34491634491634493,
      "grad_norm": 5.225584058992361,
      "learning_rate": 5.737564322469983e-05,
      "loss": 0.5044,
      "step": 670
    },
    {
      "epoch": 0.35006435006435005,
      "grad_norm": 3.1697056899798186,
      "learning_rate": 5.823327615780446e-05,
      "loss": 0.4857,
      "step": 680
    },
    {
      "epoch": 0.3552123552123552,
      "grad_norm": 3.5258824608431003,
      "learning_rate": 5.90909090909091e-05,
      "loss": 0.4747,
      "step": 690
    },
    {
      "epoch": 0.36036036036036034,
      "grad_norm": 3.5240284197467155,
      "learning_rate": 5.994854202401372e-05,
      "loss": 0.4563,
      "step": 700
    },
    {
      "epoch": 0.3655083655083655,
      "grad_norm": 3.787588207851851,
      "learning_rate": 6.080617495711835e-05,
      "loss": 0.4382,
      "step": 710
    },
    {
      "epoch": 0.37065637065637064,
      "grad_norm": 3.28493026844208,
      "learning_rate": 6.166380789022299e-05,
      "loss": 0.4216,
      "step": 720
    },
    {
      "epoch": 0.3758043758043758,
      "grad_norm": 3.144221026231138,
      "learning_rate": 6.252144082332761e-05,
      "loss": 0.417,
      "step": 730
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 2.780102841682132,
      "learning_rate": 6.337907375643225e-05,
      "loss": 0.3942,
      "step": 740
    },
    {
      "epoch": 0.3861003861003861,
      "grad_norm": 3.1996820882351713,
      "learning_rate": 6.423670668953689e-05,
      "loss": 0.3787,
      "step": 750
    },
    {
      "epoch": 0.3912483912483912,
      "grad_norm": 2.8374055946677923,
      "learning_rate": 6.50943396226415e-05,
      "loss": 0.3714,
      "step": 760
    },
    {
      "epoch": 0.3963963963963964,
      "grad_norm": 3.0094036046611454,
      "learning_rate": 6.595197255574615e-05,
      "loss": 0.3606,
      "step": 770
    },
    {
      "epoch": 0.4015444015444015,
      "grad_norm": 2.6679117500690204,
      "learning_rate": 6.680960548885078e-05,
      "loss": 0.3379,
      "step": 780
    },
    {
      "epoch": 0.4066924066924067,
      "grad_norm": 3.101181469988925,
      "learning_rate": 6.76672384219554e-05,
      "loss": 0.3271,
      "step": 790
    },
    {
      "epoch": 0.4118404118404118,
      "grad_norm": 2.421653916410696,
      "learning_rate": 6.852487135506004e-05,
      "loss": 0.3132,
      "step": 800
    },
    {
      "epoch": 0.416988416988417,
      "grad_norm": 2.1110958003741174,
      "learning_rate": 6.938250428816467e-05,
      "loss": 0.3013,
      "step": 810
    },
    {
      "epoch": 0.42213642213642216,
      "grad_norm": 1.9804288224535966,
      "learning_rate": 7.024013722126929e-05,
      "loss": 0.2892,
      "step": 820
    },
    {
      "epoch": 0.4272844272844273,
      "grad_norm": 2.0965122713743725,
      "learning_rate": 7.109777015437394e-05,
      "loss": 0.2802,
      "step": 830
    },
    {
      "epoch": 0.43243243243243246,
      "grad_norm": 2.0662383619585323,
      "learning_rate": 7.195540308747857e-05,
      "loss": 0.2666,
      "step": 840
    },
    {
      "epoch": 0.4375804375804376,
      "grad_norm": 1.9080845115386118,
      "learning_rate": 7.281303602058319e-05,
      "loss": 0.2557,
      "step": 850
    },
    {
      "epoch": 0.44272844272844275,
      "grad_norm": 2.435492621224063,
      "learning_rate": 7.367066895368783e-05,
      "loss": 0.2441,
      "step": 860
    },
    {
      "epoch": 0.44787644787644787,
      "grad_norm": 2.3478908862806707,
      "learning_rate": 7.452830188679245e-05,
      "loss": 0.2302,
      "step": 870
    },
    {
      "epoch": 0.45302445302445304,
      "grad_norm": 2.033330789035785,
      "learning_rate": 7.538593481989708e-05,
      "loss": 0.222,
      "step": 880
    },
    {
      "epoch": 0.45817245817245816,
      "grad_norm": 1.8373992771294045,
      "learning_rate": 7.624356775300173e-05,
      "loss": 0.2079,
      "step": 890
    },
    {
      "epoch": 0.46332046332046334,
      "grad_norm": 1.9518420548134319,
      "learning_rate": 7.710120068610634e-05,
      "loss": 0.1979,
      "step": 900
    },
    {
      "epoch": 0.46846846846846846,
      "grad_norm": 1.6103496909038504,
      "learning_rate": 7.795883361921097e-05,
      "loss": 0.1936,
      "step": 910
    },
    {
      "epoch": 0.47361647361647363,
      "grad_norm": 1.3694934489721085,
      "learning_rate": 7.881646655231562e-05,
      "loss": 0.1841,
      "step": 920
    },
    {
      "epoch": 0.47876447876447875,
      "grad_norm": 1.2736699951465489,
      "learning_rate": 7.967409948542024e-05,
      "loss": 0.1766,
      "step": 930
    },
    {
      "epoch": 0.4839124839124839,
      "grad_norm": 2.1829922501226444,
      "learning_rate": 8.053173241852487e-05,
      "loss": 0.1675,
      "step": 940
    },
    {
      "epoch": 0.48906048906048905,
      "grad_norm": 1.659964942426995,
      "learning_rate": 8.138936535162951e-05,
      "loss": 0.161,
      "step": 950
    },
    {
      "epoch": 0.4942084942084942,
      "grad_norm": 1.5767630911351294,
      "learning_rate": 8.224699828473413e-05,
      "loss": 0.1494,
      "step": 960
    },
    {
      "epoch": 0.49935649935649934,
      "grad_norm": 1.537782758985047,
      "learning_rate": 8.310463121783876e-05,
      "loss": 0.1445,
      "step": 970
    },
    {
      "epoch": 0.5045045045045045,
      "grad_norm": 1.604101659458472,
      "learning_rate": 8.396226415094341e-05,
      "loss": 0.1402,
      "step": 980
    },
    {
      "epoch": 0.5096525096525096,
      "grad_norm": 2.256805925155236,
      "learning_rate": 8.481989708404803e-05,
      "loss": 0.1405,
      "step": 990
    },
    {
      "epoch": 0.5148005148005148,
      "grad_norm": 1.4857578289963005,
      "learning_rate": 8.567753001715266e-05,
      "loss": 0.1321,
      "step": 1000
    },
    {
      "epoch": 0.51994851994852,
      "grad_norm": 1.0326983623178083,
      "learning_rate": 8.65351629502573e-05,
      "loss": 0.1224,
      "step": 1010
    },
    {
      "epoch": 0.525096525096525,
      "grad_norm": 1.1180976317039228,
      "learning_rate": 8.739279588336192e-05,
      "loss": 0.1189,
      "step": 1020
    },
    {
      "epoch": 0.5302445302445302,
      "grad_norm": 1.1721351234336,
      "learning_rate": 8.825042881646655e-05,
      "loss": 0.1092,
      "step": 1030
    },
    {
      "epoch": 0.5353925353925354,
      "grad_norm": 0.9640093738745075,
      "learning_rate": 8.91080617495712e-05,
      "loss": 0.1021,
      "step": 1040
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 1.0306043643976568,
      "learning_rate": 8.996569468267581e-05,
      "loss": 0.0982,
      "step": 1050
    },
    {
      "epoch": 0.5456885456885456,
      "grad_norm": 1.4999347000799717,
      "learning_rate": 9.082332761578045e-05,
      "loss": 0.0936,
      "step": 1060
    },
    {
      "epoch": 0.5508365508365508,
      "grad_norm": 1.0452281614309868,
      "learning_rate": 9.168096054888509e-05,
      "loss": 0.0878,
      "step": 1070
    },
    {
      "epoch": 0.555984555984556,
      "grad_norm": 1.2088207649167284,
      "learning_rate": 9.253859348198971e-05,
      "loss": 0.0824,
      "step": 1080
    },
    {
      "epoch": 0.5611325611325612,
      "grad_norm": 0.6373898488496081,
      "learning_rate": 9.339622641509434e-05,
      "loss": 0.0809,
      "step": 1090
    },
    {
      "epoch": 0.5662805662805663,
      "grad_norm": 0.5917955695640877,
      "learning_rate": 9.425385934819899e-05,
      "loss": 0.0742,
      "step": 1100
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.6146276278340692,
      "learning_rate": 9.51114922813036e-05,
      "loss": 0.0706,
      "step": 1110
    },
    {
      "epoch": 0.5765765765765766,
      "grad_norm": 0.8766352471472525,
      "learning_rate": 9.596912521440823e-05,
      "loss": 0.0672,
      "step": 1120
    },
    {
      "epoch": 0.5817245817245817,
      "grad_norm": 0.6678260178109867,
      "learning_rate": 9.682675814751287e-05,
      "loss": 0.0655,
      "step": 1130
    },
    {
      "epoch": 0.5868725868725869,
      "grad_norm": 0.49817357185278055,
      "learning_rate": 9.76843910806175e-05,
      "loss": 0.0611,
      "step": 1140
    },
    {
      "epoch": 0.592020592020592,
      "grad_norm": 1.0087468424307855,
      "learning_rate": 9.854202401372213e-05,
      "loss": 0.0611,
      "step": 1150
    },
    {
      "epoch": 0.5971685971685972,
      "grad_norm": 0.5362523083451194,
      "learning_rate": 9.939965694682676e-05,
      "loss": 0.0579,
      "step": 1160
    },
    {
      "epoch": 0.6023166023166023,
      "grad_norm": 0.4534299709647709,
      "learning_rate": 9.99999798272238e-05,
      "loss": 0.0549,
      "step": 1170
    },
    {
      "epoch": 0.6074646074646075,
      "grad_norm": 0.6406776231632558,
      "learning_rate": 9.999962120054416e-05,
      "loss": 0.0534,
      "step": 1180
    },
    {
      "epoch": 0.6126126126126126,
      "grad_norm": 0.5792838241902996,
      "learning_rate": 9.999881429364988e-05,
      "loss": 0.0512,
      "step": 1190
    },
    {
      "epoch": 0.6177606177606177,
      "grad_norm": 0.4275795011357426,
      "learning_rate": 9.999755911377544e-05,
      "loss": 0.0488,
      "step": 1200
    },
    {
      "epoch": 0.6229086229086229,
      "grad_norm": 0.33615545845119615,
      "learning_rate": 9.999585567217439e-05,
      "loss": 0.0476,
      "step": 1210
    },
    {
      "epoch": 0.6280566280566281,
      "grad_norm": 0.33655008372027123,
      "learning_rate": 9.99937039841192e-05,
      "loss": 0.0444,
      "step": 1220
    },
    {
      "epoch": 0.6332046332046332,
      "grad_norm": 0.43244236090431704,
      "learning_rate": 9.999110406890121e-05,
      "loss": 0.0438,
      "step": 1230
    },
    {
      "epoch": 0.6383526383526383,
      "grad_norm": 0.41139634543075815,
      "learning_rate": 9.998805594983043e-05,
      "loss": 0.042,
      "step": 1240
    },
    {
      "epoch": 0.6435006435006435,
      "grad_norm": 0.5206972896361842,
      "learning_rate": 9.998455965423525e-05,
      "loss": 0.0431,
      "step": 1250
    },
    {
      "epoch": 0.6486486486486487,
      "grad_norm": 0.3937459395109014,
      "learning_rate": 9.998061521346236e-05,
      "loss": 0.041,
      "step": 1260
    },
    {
      "epoch": 0.6537966537966537,
      "grad_norm": 0.2963678010736684,
      "learning_rate": 9.997622266287629e-05,
      "loss": 0.0393,
      "step": 1270
    },
    {
      "epoch": 0.6589446589446589,
      "grad_norm": 0.26754815540836463,
      "learning_rate": 9.997138204185923e-05,
      "loss": 0.0376,
      "step": 1280
    },
    {
      "epoch": 0.6640926640926641,
      "grad_norm": 0.29045330821553156,
      "learning_rate": 9.996609339381057e-05,
      "loss": 0.036,
      "step": 1290
    },
    {
      "epoch": 0.6692406692406693,
      "grad_norm": 0.3089797983346258,
      "learning_rate": 9.996035676614663e-05,
      "loss": 0.0356,
      "step": 1300
    },
    {
      "epoch": 0.6743886743886743,
      "grad_norm": 0.26639822964562176,
      "learning_rate": 9.99541722103001e-05,
      "loss": 0.0349,
      "step": 1310
    },
    {
      "epoch": 0.6795366795366795,
      "grad_norm": 0.45121223083554485,
      "learning_rate": 9.994753978171971e-05,
      "loss": 0.0351,
      "step": 1320
    },
    {
      "epoch": 0.6846846846846847,
      "grad_norm": 0.3667468018122777,
      "learning_rate": 9.994045953986964e-05,
      "loss": 0.0338,
      "step": 1330
    },
    {
      "epoch": 0.6898326898326899,
      "grad_norm": 0.24841154017565364,
      "learning_rate": 9.993293154822901e-05,
      "loss": 0.0328,
      "step": 1340
    },
    {
      "epoch": 0.694980694980695,
      "grad_norm": 0.283360706082952,
      "learning_rate": 9.992495587429129e-05,
      "loss": 0.0315,
      "step": 1350
    },
    {
      "epoch": 0.7001287001287001,
      "grad_norm": 0.24413900296621255,
      "learning_rate": 9.991653258956381e-05,
      "loss": 0.0311,
      "step": 1360
    },
    {
      "epoch": 0.7052767052767053,
      "grad_norm": 0.28975076041601167,
      "learning_rate": 9.990766176956697e-05,
      "loss": 0.0303,
      "step": 1370
    },
    {
      "epoch": 0.7104247104247104,
      "grad_norm": 0.26351763932473593,
      "learning_rate": 9.989834349383361e-05,
      "loss": 0.03,
      "step": 1380
    },
    {
      "epoch": 0.7155727155727156,
      "grad_norm": 0.5331124737621773,
      "learning_rate": 9.988857784590836e-05,
      "loss": 0.0291,
      "step": 1390
    },
    {
      "epoch": 0.7207207207207207,
      "grad_norm": 0.24186391898474244,
      "learning_rate": 9.98783649133468e-05,
      "loss": 0.0284,
      "step": 1400
    },
    {
      "epoch": 0.7258687258687259,
      "grad_norm": 0.21615992134341075,
      "learning_rate": 9.986770478771475e-05,
      "loss": 0.0273,
      "step": 1410
    },
    {
      "epoch": 0.731016731016731,
      "grad_norm": 0.21264572442174073,
      "learning_rate": 9.985659756458741e-05,
      "loss": 0.0267,
      "step": 1420
    },
    {
      "epoch": 0.7361647361647362,
      "grad_norm": 0.2320920564652625,
      "learning_rate": 9.984504334354847e-05,
      "loss": 0.0262,
      "step": 1430
    },
    {
      "epoch": 0.7413127413127413,
      "grad_norm": 0.2342099433653173,
      "learning_rate": 9.983304222818929e-05,
      "loss": 0.0252,
      "step": 1440
    },
    {
      "epoch": 0.7464607464607464,
      "grad_norm": 0.240834162772985,
      "learning_rate": 9.9820594326108e-05,
      "loss": 0.0245,
      "step": 1450
    },
    {
      "epoch": 0.7516087516087516,
      "grad_norm": 0.2161277200650416,
      "learning_rate": 9.980769974890833e-05,
      "loss": 0.0238,
      "step": 1460
    },
    {
      "epoch": 0.7567567567567568,
      "grad_norm": 0.24019983928841215,
      "learning_rate": 9.979435861219888e-05,
      "loss": 0.0233,
      "step": 1470
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 0.2970865898808262,
      "learning_rate": 9.978057103559188e-05,
      "loss": 0.0226,
      "step": 1480
    },
    {
      "epoch": 0.767052767052767,
      "grad_norm": 0.24780480301331292,
      "learning_rate": 9.976633714270226e-05,
      "loss": 0.0221,
      "step": 1490
    },
    {
      "epoch": 0.7722007722007722,
      "grad_norm": 0.25727613523882426,
      "learning_rate": 9.975165706114639e-05,
      "loss": 0.0214,
      "step": 1500
    },
    {
      "epoch": 0.7773487773487774,
      "grad_norm": 0.24359971967254024,
      "learning_rate": 9.973653092254113e-05,
      "loss": 0.0209,
      "step": 1510
    },
    {
      "epoch": 0.7824967824967825,
      "grad_norm": 0.2067746721639629,
      "learning_rate": 9.97209588625024e-05,
      "loss": 0.0202,
      "step": 1520
    },
    {
      "epoch": 0.7876447876447876,
      "grad_norm": 0.19304855106218685,
      "learning_rate": 9.970494102064427e-05,
      "loss": 0.0195,
      "step": 1530
    },
    {
      "epoch": 0.7927927927927928,
      "grad_norm": 0.22693160915077307,
      "learning_rate": 9.968847754057742e-05,
      "loss": 0.019,
      "step": 1540
    },
    {
      "epoch": 0.797940797940798,
      "grad_norm": 0.23060204821027103,
      "learning_rate": 9.967156856990801e-05,
      "loss": 0.0186,
      "step": 1550
    },
    {
      "epoch": 0.803088803088803,
      "grad_norm": 2.572239383211685,
      "learning_rate": 9.965421426023631e-05,
      "loss": 0.1139,
      "step": 1560
    },
    {
      "epoch": 0.8082368082368082,
      "grad_norm": 0.5437887934638203,
      "learning_rate": 9.963641476715541e-05,
      "loss": 0.0343,
      "step": 1570
    },
    {
      "epoch": 0.8133848133848134,
      "grad_norm": 0.23796250360876128,
      "learning_rate": 9.961817025024974e-05,
      "loss": 0.0235,
      "step": 1580
    },
    {
      "epoch": 0.8185328185328186,
      "grad_norm": 0.1764844020309923,
      "learning_rate": 9.959948087309362e-05,
      "loss": 0.0195,
      "step": 1590
    },
    {
      "epoch": 0.8236808236808236,
      "grad_norm": 0.17189189871306304,
      "learning_rate": 9.95803468032499e-05,
      "loss": 0.0177,
      "step": 1600
    },
    {
      "epoch": 0.8288288288288288,
      "grad_norm": 0.17751312710898323,
      "learning_rate": 9.956076821226836e-05,
      "loss": 0.0171,
      "step": 1610
    },
    {
      "epoch": 0.833976833976834,
      "grad_norm": 0.16437143945406296,
      "learning_rate": 9.954074527568426e-05,
      "loss": 0.0161,
      "step": 1620
    },
    {
      "epoch": 0.8391248391248392,
      "grad_norm": 0.15832400304822497,
      "learning_rate": 9.952027817301667e-05,
      "loss": 0.0157,
      "step": 1630
    },
    {
      "epoch": 0.8442728442728443,
      "grad_norm": 0.16117748572605023,
      "learning_rate": 9.949936708776691e-05,
      "loss": 0.0148,
      "step": 1640
    },
    {
      "epoch": 0.8494208494208494,
      "grad_norm": 0.16984391534714557,
      "learning_rate": 9.947801220741695e-05,
      "loss": 0.0143,
      "step": 1650
    },
    {
      "epoch": 0.8545688545688546,
      "grad_norm": 0.1696223065684197,
      "learning_rate": 9.945621372342762e-05,
      "loss": 0.0137,
      "step": 1660
    },
    {
      "epoch": 0.8597168597168597,
      "grad_norm": 0.1800035242900872,
      "learning_rate": 9.943397183123697e-05,
      "loss": 0.0133,
      "step": 1670
    },
    {
      "epoch": 0.8648648648648649,
      "grad_norm": 0.15249239519006993,
      "learning_rate": 9.941128673025854e-05,
      "loss": 0.0133,
      "step": 1680
    },
    {
      "epoch": 0.87001287001287,
      "grad_norm": 0.1551659031233157,
      "learning_rate": 9.938815862387953e-05,
      "loss": 0.0132,
      "step": 1690
    },
    {
      "epoch": 0.8751608751608752,
      "grad_norm": 0.14348975402940561,
      "learning_rate": 9.936458771945891e-05,
      "loss": 0.0131,
      "step": 1700
    },
    {
      "epoch": 0.8803088803088803,
      "grad_norm": 0.140601933336413,
      "learning_rate": 9.934057422832572e-05,
      "loss": 0.0123,
      "step": 1710
    },
    {
      "epoch": 0.8854568854568855,
      "grad_norm": 0.1275666846983677,
      "learning_rate": 9.931611836577703e-05,
      "loss": 0.012,
      "step": 1720
    },
    {
      "epoch": 0.8906048906048906,
      "grad_norm": 0.13327069275308084,
      "learning_rate": 9.92912203510761e-05,
      "loss": 0.012,
      "step": 1730
    },
    {
      "epoch": 0.8957528957528957,
      "grad_norm": 0.1658754837807317,
      "learning_rate": 9.926588040745035e-05,
      "loss": 0.0117,
      "step": 1740
    },
    {
      "epoch": 0.9009009009009009,
      "grad_norm": 0.14050126640493185,
      "learning_rate": 9.924009876208944e-05,
      "loss": 0.0116,
      "step": 1750
    },
    {
      "epoch": 0.9060489060489061,
      "grad_norm": 0.1413904681709387,
      "learning_rate": 9.921387564614313e-05,
      "loss": 0.0111,
      "step": 1760
    },
    {
      "epoch": 0.9111969111969112,
      "grad_norm": 0.13671787855200412,
      "learning_rate": 9.918721129471927e-05,
      "loss": 0.0113,
      "step": 1770
    },
    {
      "epoch": 0.9163449163449163,
      "grad_norm": 0.1300287060008007,
      "learning_rate": 9.916010594688173e-05,
      "loss": 0.011,
      "step": 1780
    },
    {
      "epoch": 0.9214929214929215,
      "grad_norm": 0.12272142768731135,
      "learning_rate": 9.913255984564814e-05,
      "loss": 0.0104,
      "step": 1790
    },
    {
      "epoch": 0.9266409266409267,
      "grad_norm": 0.13585333911080333,
      "learning_rate": 9.910457323798783e-05,
      "loss": 0.0101,
      "step": 1800
    }
  ],
  "logging_steps": 10,
  "max_steps": 11658,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 1800,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 161574079168512.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}