{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9994342824816141,
  "eval_steps": 500,
  "global_step": 1325,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0007542900245144258,
      "grad_norm": 24.0743322267596,
      "learning_rate": 7.518796992481203e-08,
      "loss": 1.4329,
      "step": 1
    },
    {
      "epoch": 0.003771450122572129,
      "grad_norm": 24.762983495374982,
      "learning_rate": 3.7593984962406015e-07,
      "loss": 1.417,
      "step": 5
    },
    {
      "epoch": 0.007542900245144258,
      "grad_norm": 8.686384748409985,
      "learning_rate": 7.518796992481203e-07,
      "loss": 1.3206,
      "step": 10
    },
    {
      "epoch": 0.011314350367716388,
      "grad_norm": 10.36105058434612,
      "learning_rate": 1.1278195488721805e-06,
      "loss": 1.1564,
      "step": 15
    },
    {
      "epoch": 0.015085800490288516,
      "grad_norm": 3.2096721215381483,
      "learning_rate": 1.5037593984962406e-06,
      "loss": 1.0254,
      "step": 20
    },
    {
      "epoch": 0.018857250612860643,
      "grad_norm": 2.8041993817362436,
      "learning_rate": 1.8796992481203007e-06,
      "loss": 0.9632,
      "step": 25
    },
    {
      "epoch": 0.022628700735432775,
      "grad_norm": 2.4316167661473758,
      "learning_rate": 2.255639097744361e-06,
      "loss": 0.9285,
      "step": 30
    },
    {
      "epoch": 0.026400150858004904,
      "grad_norm": 2.347169321112029,
      "learning_rate": 2.631578947368421e-06,
      "loss": 0.9228,
      "step": 35
    },
    {
      "epoch": 0.030171600980577033,
      "grad_norm": 2.23082557930139,
      "learning_rate": 3.007518796992481e-06,
      "loss": 0.8955,
      "step": 40
    },
    {
      "epoch": 0.03394305110314916,
      "grad_norm": 2.3187330629715484,
      "learning_rate": 3.3834586466165413e-06,
      "loss": 0.9049,
      "step": 45
    },
    {
      "epoch": 0.037714501225721286,
      "grad_norm": 2.239439563631581,
      "learning_rate": 3.7593984962406014e-06,
      "loss": 0.894,
      "step": 50
    },
    {
      "epoch": 0.04148595134829342,
      "grad_norm": 2.3025921470010267,
      "learning_rate": 4.135338345864662e-06,
      "loss": 0.8807,
      "step": 55
    },
    {
      "epoch": 0.04525740147086555,
      "grad_norm": 2.387099716939349,
      "learning_rate": 4.511278195488722e-06,
      "loss": 0.875,
      "step": 60
    },
    {
      "epoch": 0.04902885159343768,
      "grad_norm": 2.3813541654085046,
      "learning_rate": 4.887218045112782e-06,
      "loss": 0.8674,
      "step": 65
    },
    {
      "epoch": 0.05280030171600981,
      "grad_norm": 2.609644080569622,
      "learning_rate": 5.263157894736842e-06,
      "loss": 0.8655,
      "step": 70
    },
    {
      "epoch": 0.056571751838581937,
      "grad_norm": 2.571607734979684,
      "learning_rate": 5.639097744360903e-06,
      "loss": 0.862,
      "step": 75
    },
    {
      "epoch": 0.060343201961154065,
      "grad_norm": 2.2582812119753903,
      "learning_rate": 6.015037593984962e-06,
      "loss": 0.8518,
      "step": 80
    },
    {
      "epoch": 0.06411465208372619,
      "grad_norm": 2.450433458259798,
      "learning_rate": 6.390977443609023e-06,
      "loss": 0.856,
      "step": 85
    },
    {
      "epoch": 0.06788610220629832,
      "grad_norm": 2.58014673334683,
      "learning_rate": 6.766917293233083e-06,
      "loss": 0.8528,
      "step": 90
    },
    {
      "epoch": 0.07165755232887044,
      "grad_norm": 2.3740208001929743,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 0.8331,
      "step": 95
    },
    {
      "epoch": 0.07542900245144257,
      "grad_norm": 2.413650287157212,
      "learning_rate": 7.518796992481203e-06,
      "loss": 0.8605,
      "step": 100
    },
    {
      "epoch": 0.07920045257401472,
      "grad_norm": 2.355976041456668,
      "learning_rate": 7.894736842105265e-06,
      "loss": 0.8389,
      "step": 105
    },
    {
      "epoch": 0.08297190269658684,
      "grad_norm": 2.3146510995724334,
      "learning_rate": 8.270676691729324e-06,
      "loss": 0.8398,
      "step": 110
    },
    {
      "epoch": 0.08674335281915897,
      "grad_norm": 2.539423496991191,
      "learning_rate": 8.646616541353385e-06,
      "loss": 0.8281,
      "step": 115
    },
    {
      "epoch": 0.0905148029417311,
      "grad_norm": 2.3110234365436115,
      "learning_rate": 9.022556390977444e-06,
      "loss": 0.8355,
      "step": 120
    },
    {
      "epoch": 0.09428625306430323,
      "grad_norm": 2.3074854008707644,
      "learning_rate": 9.398496240601505e-06,
      "loss": 0.8378,
      "step": 125
    },
    {
      "epoch": 0.09805770318687536,
      "grad_norm": 2.198478602649139,
      "learning_rate": 9.774436090225564e-06,
      "loss": 0.8255,
      "step": 130
    },
    {
      "epoch": 0.10182915330944749,
      "grad_norm": 2.372634478935117,
      "learning_rate": 9.999930538168684e-06,
      "loss": 0.8152,
      "step": 135
    },
    {
      "epoch": 0.10560060343201962,
      "grad_norm": 2.3429810617684406,
      "learning_rate": 9.999149114730812e-06,
      "loss": 0.829,
      "step": 140
    },
    {
      "epoch": 0.10937205355459174,
      "grad_norm": 2.2344447897248783,
      "learning_rate": 9.997499576714369e-06,
      "loss": 0.8299,
      "step": 145
    },
    {
      "epoch": 0.11314350367716387,
      "grad_norm": 2.241744206103708,
      "learning_rate": 9.9949822105657e-06,
      "loss": 0.8253,
      "step": 150
    },
    {
      "epoch": 0.116914953799736,
      "grad_norm": 2.2980391560118862,
      "learning_rate": 9.991597453431652e-06,
      "loss": 0.8258,
      "step": 155
    },
    {
      "epoch": 0.12068640392230813,
      "grad_norm": 2.501411375935894,
      "learning_rate": 9.987345893083652e-06,
      "loss": 0.8165,
      "step": 160
    },
    {
      "epoch": 0.12445785404488026,
      "grad_norm": 2.286238996613632,
      "learning_rate": 9.982228267815644e-06,
      "loss": 0.8128,
      "step": 165
    },
    {
      "epoch": 0.12822930416745237,
      "grad_norm": 2.179417789665571,
      "learning_rate": 9.97624546631588e-06,
      "loss": 0.8193,
      "step": 170
    },
    {
      "epoch": 0.1320007542900245,
      "grad_norm": 2.186299499644603,
      "learning_rate": 9.969398527512607e-06,
      "loss": 0.8153,
      "step": 175
    },
    {
      "epoch": 0.13577220441259663,
      "grad_norm": 2.177003107850399,
      "learning_rate": 9.96168864039364e-06,
      "loss": 0.8004,
      "step": 180
    },
    {
      "epoch": 0.13954365453516876,
      "grad_norm": 2.5007296087774575,
      "learning_rate": 9.9531171437999e-06,
      "loss": 0.8057,
      "step": 185
    },
    {
      "epoch": 0.1433151046577409,
      "grad_norm": 2.2265359668215696,
      "learning_rate": 9.943685526192921e-06,
      "loss": 0.8045,
      "step": 190
    },
    {
      "epoch": 0.14708655478031302,
      "grad_norm": 2.1009649317490116,
      "learning_rate": 9.93339542539637e-06,
      "loss": 0.7964,
      "step": 195
    },
    {
      "epoch": 0.15085800490288515,
      "grad_norm": 2.0704593465422008,
      "learning_rate": 9.922248628311641e-06,
      "loss": 0.8007,
      "step": 200
    },
    {
      "epoch": 0.1546294550254573,
      "grad_norm": 2.2278475317912694,
      "learning_rate": 9.91024707060755e-06,
      "loss": 0.8074,
      "step": 205
    },
    {
      "epoch": 0.15840090514802943,
      "grad_norm": 2.3927523335337293,
      "learning_rate": 9.89739283638421e-06,
      "loss": 0.7964,
      "step": 210
    },
    {
      "epoch": 0.16217235527060156,
      "grad_norm": 2.2597054171285826,
      "learning_rate": 9.883688157811102e-06,
      "loss": 0.7925,
      "step": 215
    },
    {
      "epoch": 0.1659438053931737,
      "grad_norm": 2.2971840831123806,
      "learning_rate": 9.86913541473948e-06,
      "loss": 0.7928,
      "step": 220
    },
    {
      "epoch": 0.16971525551574582,
      "grad_norm": 2.161562265562174,
      "learning_rate": 9.853737134289086e-06,
      "loss": 0.78,
      "step": 225
    },
    {
      "epoch": 0.17348670563831795,
      "grad_norm": 2.087919478225615,
      "learning_rate": 9.83749599040931e-06,
      "loss": 0.7852,
      "step": 230
    },
    {
      "epoch": 0.17725815576089007,
      "grad_norm": 2.1464178337018422,
      "learning_rate": 9.820414803414862e-06,
      "loss": 0.7864,
      "step": 235
    },
    {
      "epoch": 0.1810296058834622,
      "grad_norm": 2.1956780101934115,
      "learning_rate": 9.802496539496008e-06,
      "loss": 0.777,
      "step": 240
    },
    {
      "epoch": 0.18480105600603433,
      "grad_norm": 2.2171827211629456,
      "learning_rate": 9.783744310203492e-06,
      "loss": 0.7738,
      "step": 245
    },
    {
      "epoch": 0.18857250612860646,
      "grad_norm": 2.1354704419052,
      "learning_rate": 9.764161371908195e-06,
      "loss": 0.7833,
      "step": 250
    },
    {
      "epoch": 0.1923439562511786,
      "grad_norm": 2.244895598639118,
      "learning_rate": 9.743751125235674e-06,
      "loss": 0.7832,
      "step": 255
    },
    {
      "epoch": 0.19611540637375072,
      "grad_norm": 2.1690612043700983,
      "learning_rate": 9.722517114475622e-06,
      "loss": 0.7759,
      "step": 260
    },
    {
      "epoch": 0.19988685649632285,
      "grad_norm": 2.465068879237022,
      "learning_rate": 9.7004630269664e-06,
      "loss": 0.7736,
      "step": 265
    },
    {
      "epoch": 0.20365830661889497,
      "grad_norm": 2.3445933624332973,
      "learning_rate": 9.67759269245472e-06,
      "loss": 0.7617,
      "step": 270
    },
    {
      "epoch": 0.2074297567414671,
      "grad_norm": 2.131498314204489,
      "learning_rate": 9.6539100824306e-06,
      "loss": 0.7616,
      "step": 275
    },
    {
      "epoch": 0.21120120686403923,
      "grad_norm": 2.365920326550214,
      "learning_rate": 9.629419309437704e-06,
      "loss": 0.7555,
      "step": 280
    },
    {
      "epoch": 0.21497265698661136,
      "grad_norm": 2.330336022690634,
      "learning_rate": 9.60412462635919e-06,
      "loss": 0.7541,
      "step": 285
    },
    {
      "epoch": 0.2187441071091835,
      "grad_norm": 2.2560710591306634,
      "learning_rate": 9.578030425679188e-06,
      "loss": 0.751,
      "step": 290
    },
    {
      "epoch": 0.22251555723175562,
      "grad_norm": 2.054455615320725,
      "learning_rate": 9.551141238720027e-06,
      "loss": 0.7491,
      "step": 295
    },
    {
      "epoch": 0.22628700735432775,
      "grad_norm": 2.3326134119870465,
      "learning_rate": 9.523461734855373e-06,
      "loss": 0.7494,
      "step": 300
    },
    {
      "epoch": 0.23005845747689987,
      "grad_norm": 2.0451775689136995,
      "learning_rate": 9.494996720699363e-06,
      "loss": 0.7527,
      "step": 305
    },
    {
      "epoch": 0.233829907599472,
      "grad_norm": 2.2108967714871275,
      "learning_rate": 9.465751139271943e-06,
      "loss": 0.751,
      "step": 310
    },
    {
      "epoch": 0.23760135772204413,
      "grad_norm": 2.09565099808324,
      "learning_rate": 9.435730069140486e-06,
      "loss": 0.741,
      "step": 315
    },
    {
      "epoch": 0.24137280784461626,
      "grad_norm": 2.1588491738436937,
      "learning_rate": 9.404938723537894e-06,
      "loss": 0.7429,
      "step": 320
    },
    {
      "epoch": 0.2451442579671884,
      "grad_norm": 2.056502617886576,
      "learning_rate": 9.373382449457305e-06,
      "loss": 0.7341,
      "step": 325
    },
    {
      "epoch": 0.24891570808976052,
      "grad_norm": 2.067525516090952,
      "learning_rate": 9.341066726723574e-06,
      "loss": 0.7522,
      "step": 330
    },
    {
      "epoch": 0.2526871582123326,
      "grad_norm": 2.141985494588096,
      "learning_rate": 9.307997167041688e-06,
      "loss": 0.736,
      "step": 335
    },
    {
      "epoch": 0.25645860833490475,
      "grad_norm": 2.175709280838213,
      "learning_rate": 9.274179513022284e-06,
      "loss": 0.7427,
      "step": 340
    },
    {
      "epoch": 0.2602300584574769,
      "grad_norm": 2.661524805942838,
      "learning_rate": 9.23961963718442e-06,
      "loss": 0.7354,
      "step": 345
    },
    {
      "epoch": 0.264001508580049,
      "grad_norm": 2.3806741154817606,
      "learning_rate": 9.204323540935808e-06,
      "loss": 0.7171,
      "step": 350
    },
    {
      "epoch": 0.26777295870262113,
      "grad_norm": 2.2844034910230384,
      "learning_rate": 9.168297353530654e-06,
      "loss": 0.7246,
      "step": 355
    },
    {
      "epoch": 0.27154440882519326,
      "grad_norm": 2.2290764994416756,
      "learning_rate": 9.131547331005288e-06,
      "loss": 0.7219,
      "step": 360
    },
    {
      "epoch": 0.2753158589477654,
      "grad_norm": 2.1718562436320727,
      "learning_rate": 9.094079855091797e-06,
      "loss": 0.7106,
      "step": 365
    },
    {
      "epoch": 0.2790873090703375,
      "grad_norm": 2.055976595872544,
      "learning_rate": 9.055901432109821e-06,
      "loss": 0.7242,
      "step": 370
    },
    {
      "epoch": 0.28285875919290965,
      "grad_norm": 2.2447216963743575,
      "learning_rate": 9.017018691836714e-06,
      "loss": 0.7269,
      "step": 375
    },
    {
      "epoch": 0.2866302093154818,
      "grad_norm": 2.108026977341414,
      "learning_rate": 8.97743838635626e-06,
      "loss": 0.7253,
      "step": 380
    },
    {
      "epoch": 0.2904016594380539,
      "grad_norm": 2.149969361483023,
      "learning_rate": 8.937167388886163e-06,
      "loss": 0.7154,
      "step": 385
    },
    {
      "epoch": 0.29417310956062603,
      "grad_norm": 2.1649261230436543,
      "learning_rate": 8.896212692584503e-06,
      "loss": 0.7215,
      "step": 390
    },
    {
      "epoch": 0.29794455968319816,
      "grad_norm": 1.9995253010334968,
      "learning_rate": 8.854581409335341e-06,
      "loss": 0.7148,
      "step": 395
    },
    {
      "epoch": 0.3017160098057703,
      "grad_norm": 2.2916154939190783,
      "learning_rate": 8.812280768513732e-06,
      "loss": 0.7089,
      "step": 400
    },
    {
      "epoch": 0.3054874599283425,
      "grad_norm": 2.1141628178573466,
      "learning_rate": 8.76931811573033e-06,
      "loss": 0.708,
      "step": 405
    },
    {
      "epoch": 0.3092589100509146,
      "grad_norm": 2.175109844911728,
      "learning_rate": 8.725700911555792e-06,
      "loss": 0.7037,
      "step": 410
    },
    {
      "epoch": 0.31303036017348673,
      "grad_norm": 2.101023552691967,
      "learning_rate": 8.681436730225245e-06,
      "loss": 0.7053,
      "step": 415
    },
    {
      "epoch": 0.31680181029605886,
      "grad_norm": 2.101523445074263,
      "learning_rate": 8.636533258322995e-06,
      "loss": 0.6861,
      "step": 420
    },
    {
      "epoch": 0.320573260418631,
      "grad_norm": 5.819641297709661,
      "learning_rate": 8.590998293447728e-06,
      "loss": 0.6921,
      "step": 425
    },
    {
      "epoch": 0.3243447105412031,
      "grad_norm": 2.3399412061978784,
      "learning_rate": 8.544839742858449e-06,
      "loss": 0.691,
      "step": 430
    },
    {
      "epoch": 0.32811616066377525,
      "grad_norm": 2.2092909626343173,
      "learning_rate": 8.49806562210136e-06,
      "loss": 0.6934,
      "step": 435
    },
    {
      "epoch": 0.3318876107863474,
      "grad_norm": 2.0855142628275374,
      "learning_rate": 8.450684053617936e-06,
      "loss": 0.693,
      "step": 440
    },
    {
      "epoch": 0.3356590609089195,
      "grad_norm": 2.259615514364146,
      "learning_rate": 8.402703265334455e-06,
      "loss": 0.6949,
      "step": 445
    },
    {
      "epoch": 0.33943051103149163,
      "grad_norm": 2.176855477838487,
      "learning_rate": 8.35413158923319e-06,
      "loss": 0.6943,
      "step": 450
    },
    {
      "epoch": 0.34320196115406376,
      "grad_norm": 5.3013314444050295,
      "learning_rate": 8.304977459905539e-06,
      "loss": 0.693,
      "step": 455
    },
    {
      "epoch": 0.3469734112766359,
      "grad_norm": 2.3136109012385853,
      "learning_rate": 8.255249413087342e-06,
      "loss": 0.6839,
      "step": 460
    },
    {
      "epoch": 0.350744861399208,
      "grad_norm": 2.5530506882002775,
      "learning_rate": 8.20495608417663e-06,
      "loss": 0.6837,
      "step": 465
    },
    {
      "epoch": 0.35451631152178015,
      "grad_norm": 2.2039921142230923,
      "learning_rate": 8.154106206734056e-06,
      "loss": 0.6757,
      "step": 470
    },
    {
      "epoch": 0.3582877616443523,
      "grad_norm": 2.0873079113549005,
      "learning_rate": 8.102708610966306e-06,
      "loss": 0.6845,
      "step": 475
    },
    {
      "epoch": 0.3620592117669244,
      "grad_norm": 2.2652146917109994,
      "learning_rate": 8.050772222192703e-06,
      "loss": 0.679,
      "step": 480
    },
    {
      "epoch": 0.36583066188949653,
      "grad_norm": 2.342496143877203,
      "learning_rate": 7.998306059295302e-06,
      "loss": 0.6763,
      "step": 485
    },
    {
      "epoch": 0.36960211201206866,
      "grad_norm": 2.1611026997479246,
      "learning_rate": 7.945319233152748e-06,
      "loss": 0.6731,
      "step": 490
    },
    {
      "epoch": 0.3733735621346408,
      "grad_norm": 2.1522125366560556,
      "learning_rate": 7.891820945058137e-06,
      "loss": 0.6704,
      "step": 495
    },
    {
      "epoch": 0.3771450122572129,
      "grad_norm": 2.225656595731522,
      "learning_rate": 7.837820485121196e-06,
      "loss": 0.6577,
      "step": 500
    },
    {
      "epoch": 0.38091646237978505,
      "grad_norm": 2.05947893221226,
      "learning_rate": 7.783327230655036e-06,
      "loss": 0.657,
      "step": 505
    },
    {
      "epoch": 0.3846879125023572,
      "grad_norm": 2.0489553480122553,
      "learning_rate": 7.728350644547748e-06,
      "loss": 0.6599,
      "step": 510
    },
    {
      "epoch": 0.3884593626249293,
      "grad_norm": 2.3536125168662134,
      "learning_rate": 7.672900273619158e-06,
      "loss": 0.6379,
      "step": 515
    },
    {
      "epoch": 0.39223081274750143,
      "grad_norm": 2.278757928608761,
      "learning_rate": 7.616985746962995e-06,
      "loss": 0.6584,
      "step": 520
    },
    {
      "epoch": 0.39600226287007356,
      "grad_norm": 2.2890835012690385,
      "learning_rate": 7.560616774274775e-06,
      "loss": 0.6675,
      "step": 525
    },
    {
      "epoch": 0.3997737129926457,
      "grad_norm": 2.052679763492931,
      "learning_rate": 7.503803144165691e-06,
      "loss": 0.6532,
      "step": 530
    },
    {
      "epoch": 0.4035451631152178,
      "grad_norm": 2.1962590249997804,
      "learning_rate": 7.446554722462789e-06,
      "loss": 0.6452,
      "step": 535
    },
    {
      "epoch": 0.40731661323778995,
      "grad_norm": 2.1492418404505513,
      "learning_rate": 7.388881450495753e-06,
      "loss": 0.6399,
      "step": 540
    },
    {
      "epoch": 0.4110880633603621,
      "grad_norm": 2.140785983967721,
      "learning_rate": 7.33079334337056e-06,
      "loss": 0.6519,
      "step": 545
    },
    {
      "epoch": 0.4148595134829342,
      "grad_norm": 2.2023514849237857,
      "learning_rate": 7.272300488230342e-06,
      "loss": 0.6451,
      "step": 550
    },
    {
      "epoch": 0.41863096360550633,
      "grad_norm": 2.1139769229137197,
      "learning_rate": 7.213413042503714e-06,
      "loss": 0.6465,
      "step": 555
    },
    {
      "epoch": 0.42240241372807846,
      "grad_norm": 2.1321527320416216,
      "learning_rate": 7.154141232140925e-06,
      "loss": 0.632,
      "step": 560
    },
    {
      "epoch": 0.4261738638506506,
      "grad_norm": 2.1408532489827383,
      "learning_rate": 7.094495349838093e-06,
      "loss": 0.64,
      "step": 565
    },
    {
      "epoch": 0.4299453139732227,
      "grad_norm": 2.1424271774917343,
      "learning_rate": 7.034485753249846e-06,
      "loss": 0.6318,
      "step": 570
    },
    {
      "epoch": 0.43371676409579485,
      "grad_norm": 2.4132135353951223,
      "learning_rate": 6.9741228631907e-06,
      "loss": 0.6388,
      "step": 575
    },
    {
      "epoch": 0.437488214218367,
      "grad_norm": 2.2228430360361373,
      "learning_rate": 6.913417161825449e-06,
      "loss": 0.6232,
      "step": 580
    },
    {
      "epoch": 0.4412596643409391,
      "grad_norm": 2.101068755168653,
      "learning_rate": 6.852379190848923e-06,
      "loss": 0.6274,
      "step": 585
    },
    {
      "epoch": 0.44503111446351123,
      "grad_norm": 2.283059476668117,
      "learning_rate": 6.791019549655384e-06,
      "loss": 0.6178,
      "step": 590
    },
    {
      "epoch": 0.44880256458608336,
      "grad_norm": 2.1112204908741137,
      "learning_rate": 6.72934889349792e-06,
      "loss": 0.631,
      "step": 595
    },
    {
      "epoch": 0.4525740147086555,
      "grad_norm": 2.1049933744197813,
      "learning_rate": 6.667377931638144e-06,
      "loss": 0.6134,
      "step": 600
    },
    {
      "epoch": 0.4563454648312276,
      "grad_norm": 2.235411557732047,
      "learning_rate": 6.605117425486483e-06,
      "loss": 0.6119,
      "step": 605
    },
    {
      "epoch": 0.46011691495379975,
      "grad_norm": 2.0415112264754107,
      "learning_rate": 6.542578186733453e-06,
      "loss": 0.6173,
      "step": 610
    },
    {
      "epoch": 0.4638883650763719,
      "grad_norm": 2.185203312850119,
      "learning_rate": 6.47977107547217e-06,
      "loss": 0.6219,
      "step": 615
    },
    {
      "epoch": 0.467659815198944,
      "grad_norm": 2.21664954602057,
      "learning_rate": 6.4167069983124765e-06,
      "loss": 0.6144,
      "step": 620
    },
    {
      "epoch": 0.47143126532151614,
      "grad_norm": 2.1363794732205807,
      "learning_rate": 6.353396906486971e-06,
      "loss": 0.621,
      "step": 625
    },
    {
      "epoch": 0.47520271544408826,
      "grad_norm": 2.2099254913643502,
      "learning_rate": 6.289851793949311e-06,
      "loss": 0.6118,
      "step": 630
    },
    {
      "epoch": 0.4789741655666604,
      "grad_norm": 2.3004798131436055,
      "learning_rate": 6.2260826954650795e-06,
      "loss": 0.6197,
      "step": 635
    },
    {
      "epoch": 0.4827456156892325,
      "grad_norm": 2.154968570593537,
      "learning_rate": 6.1621006846955676e-06,
      "loss": 0.6012,
      "step": 640
    },
    {
      "epoch": 0.48651706581180465,
      "grad_norm": 2.2605674779702847,
      "learning_rate": 6.097916872274815e-06,
      "loss": 0.601,
      "step": 645
    },
    {
      "epoch": 0.4902885159343768,
      "grad_norm": 2.2052321176133134,
      "learning_rate": 6.033542403880216e-06,
      "loss": 0.6092,
      "step": 650
    },
    {
      "epoch": 0.4940599660569489,
      "grad_norm": 2.1808520045523427,
      "learning_rate": 5.96898845829704e-06,
      "loss": 0.6059,
      "step": 655
    },
    {
      "epoch": 0.49783141617952104,
      "grad_norm": 2.300580579687932,
      "learning_rate": 5.904266245477219e-06,
      "loss": 0.5946,
      "step": 660
    },
    {
      "epoch": 0.5016028663020932,
      "grad_norm": 2.0297693933559398,
      "learning_rate": 5.839387004592705e-06,
      "loss": 0.5859,
      "step": 665
    },
    {
      "epoch": 0.5053743164246652,
      "grad_norm": 2.0914074101770623,
      "learning_rate": 5.7743620020837565e-06,
      "loss": 0.598,
      "step": 670
    },
    {
      "epoch": 0.5091457665472374,
      "grad_norm": 2.0766975702892836,
      "learning_rate": 5.709202529702501e-06,
      "loss": 0.5879,
      "step": 675
    },
    {
      "epoch": 0.5129172166698095,
      "grad_norm": 2.213588628181455,
      "learning_rate": 5.643919902552091e-06,
      "loss": 0.5913,
      "step": 680
    },
    {
      "epoch": 0.5166886667923817,
      "grad_norm": 2.1397173162628156,
      "learning_rate": 5.578525457121807e-06,
      "loss": 0.583,
      "step": 685
    },
    {
      "epoch": 0.5204601169149538,
      "grad_norm": 2.2408994529326103,
      "learning_rate": 5.513030549318453e-06,
      "loss": 0.5783,
      "step": 690
    },
    {
      "epoch": 0.5242315670375259,
      "grad_norm": 2.243980907997254,
      "learning_rate": 5.447446552494369e-06,
      "loss": 0.5818,
      "step": 695
    },
    {
      "epoch": 0.528003017160098,
      "grad_norm": 2.0846263742726165,
      "learning_rate": 5.381784855472424e-06,
      "loss": 0.5843,
      "step": 700
    },
    {
      "epoch": 0.5317744672826702,
      "grad_norm": 2.1221104917928284,
      "learning_rate": 5.316056860568318e-06,
      "loss": 0.5921,
      "step": 705
    },
    {
      "epoch": 0.5355459174052423,
      "grad_norm": 2.12080931648079,
      "learning_rate": 5.250273981610543e-06,
      "loss": 0.5782,
      "step": 710
    },
    {
      "epoch": 0.5393173675278145,
      "grad_norm": 2.1035292957689036,
      "learning_rate": 5.184447641958339e-06,
      "loss": 0.5822,
      "step": 715
    },
    {
      "epoch": 0.5430888176503865,
      "grad_norm": 2.06196021971329,
      "learning_rate": 5.118589272518008e-06,
      "loss": 0.5728,
      "step": 720
    },
    {
      "epoch": 0.5468602677729587,
      "grad_norm": 2.055521224880793,
      "learning_rate": 5.052710309757899e-06,
      "loss": 0.5879,
      "step": 725
    },
    {
      "epoch": 0.5506317178955308,
      "grad_norm": 2.1354094712615774,
      "learning_rate": 4.9868221937224435e-06,
      "loss": 0.5767,
      "step": 730
    },
    {
      "epoch": 0.554403168018103,
      "grad_norm": 2.152252412903219,
      "learning_rate": 4.920936366045564e-06,
      "loss": 0.5769,
      "step": 735
    },
    {
      "epoch": 0.558174618140675,
      "grad_norm": 2.2029723753911066,
      "learning_rate": 4.8550642679637985e-06,
      "loss": 0.578,
      "step": 740
    },
    {
      "epoch": 0.5619460682632472,
      "grad_norm": 2.129220459700306,
      "learning_rate": 4.789217338329515e-06,
      "loss": 0.5599,
      "step": 745
    },
    {
      "epoch": 0.5657175183858193,
      "grad_norm": 2.3375640799042507,
      "learning_rate": 4.723407011624509e-06,
      "loss": 0.5624,
      "step": 750
    },
    {
      "epoch": 0.5694889685083915,
      "grad_norm": 2.086098519936994,
      "learning_rate": 4.657644715974397e-06,
      "loss": 0.5594,
      "step": 755
    },
    {
      "epoch": 0.5732604186309636,
      "grad_norm": 2.0842869926932672,
      "learning_rate": 4.591941871164075e-06,
      "loss": 0.5705,
      "step": 760
    },
    {
      "epoch": 0.5770318687535357,
      "grad_norm": 2.266813054356403,
      "learning_rate": 4.526309886654659e-06,
      "loss": 0.554,
      "step": 765
    },
    {
      "epoch": 0.5808033188761078,
      "grad_norm": 2.1089635718005333,
      "learning_rate": 4.4607601596021995e-06,
      "loss": 0.5583,
      "step": 770
    },
    {
      "epoch": 0.58457476899868,
      "grad_norm": 2.2148767655660984,
      "learning_rate": 4.395304072878533e-06,
      "loss": 0.5584,
      "step": 775
    },
    {
      "epoch": 0.5883462191212521,
      "grad_norm": 2.0165033312706164,
      "learning_rate": 4.329952993094627e-06,
      "loss": 0.5587,
      "step": 780
    },
    {
      "epoch": 0.5921176692438243,
      "grad_norm": 2.0613406319368206,
      "learning_rate": 4.264718268626729e-06,
      "loss": 0.5549,
      "step": 785
    },
    {
      "epoch": 0.5958891193663963,
      "grad_norm": 2.23703436211283,
      "learning_rate": 4.199611227645705e-06,
      "loss": 0.5608,
      "step": 790
    },
    {
      "epoch": 0.5996605694889685,
      "grad_norm": 2.1229182241921354,
      "learning_rate": 4.134643176149861e-06,
      "loss": 0.5423,
      "step": 795
    },
    {
      "epoch": 0.6034320196115406,
      "grad_norm": 2.16174398622042,
      "learning_rate": 4.069825396001643e-06,
      "loss": 0.5496,
      "step": 800
    },
    {
      "epoch": 0.6072034697341128,
      "grad_norm": 2.128117272383355,
      "learning_rate": 4.005169142968503e-06,
      "loss": 0.5519,
      "step": 805
    },
    {
      "epoch": 0.610974919856685,
      "grad_norm": 2.0175251911767216,
      "learning_rate": 3.94068564476831e-06,
      "loss": 0.5475,
      "step": 810
    },
    {
      "epoch": 0.614746369979257,
      "grad_norm": 2.0746720466907558,
      "learning_rate": 3.876386099119635e-06,
      "loss": 0.5461,
      "step": 815
    },
    {
      "epoch": 0.6185178201018292,
      "grad_norm": 2.130560812479597,
      "learning_rate": 3.8122816717972287e-06,
      "loss": 0.5407,
      "step": 820
    },
    {
      "epoch": 0.6222892702244013,
      "grad_norm": 2.1146940344656375,
      "learning_rate": 3.7483834946930682e-06,
      "loss": 0.5504,
      "step": 825
    },
    {
      "epoch": 0.6260607203469735,
      "grad_norm": 2.174000452771408,
      "learning_rate": 3.684702663883266e-06,
      "loss": 0.5497,
      "step": 830
    },
    {
      "epoch": 0.6298321704695455,
      "grad_norm": 2.173713479741331,
      "learning_rate": 3.6212502377012164e-06,
      "loss": 0.5456,
      "step": 835
    },
    {
      "epoch": 0.6336036205921177,
      "grad_norm": 2.1517414568189275,
      "learning_rate": 3.558037234817291e-06,
      "loss": 0.5344,
      "step": 840
    },
    {
      "epoch": 0.6373750707146898,
      "grad_norm": 2.0901545430594837,
      "learning_rate": 3.495074632325407e-06,
      "loss": 0.5357,
      "step": 845
    },
    {
      "epoch": 0.641146520837262,
      "grad_norm": 2.1721190985604073,
      "learning_rate": 3.4323733638368427e-06,
      "loss": 0.5366,
      "step": 850
    },
    {
      "epoch": 0.644917970959834,
      "grad_norm": 2.1159591065311836,
      "learning_rate": 3.3699443175815773e-06,
      "loss": 0.5209,
      "step": 855
    },
    {
      "epoch": 0.6486894210824062,
      "grad_norm": 2.092825596814058,
      "learning_rate": 3.3077983345175295e-06,
      "loss": 0.5357,
      "step": 860
    },
    {
      "epoch": 0.6524608712049783,
      "grad_norm": 2.112302580054298,
      "learning_rate": 3.2459462064479972e-06,
      "loss": 0.5399,
      "step": 865
    },
    {
      "epoch": 0.6562323213275505,
      "grad_norm": 2.118173020663562,
      "learning_rate": 3.1843986741476285e-06,
      "loss": 0.5381,
      "step": 870
    },
    {
      "epoch": 0.6600037714501226,
      "grad_norm": 2.089034293925172,
      "learning_rate": 3.1231664254972736e-06,
      "loss": 0.5239,
      "step": 875
    },
    {
      "epoch": 0.6637752215726948,
      "grad_norm": 2.129522823953631,
      "learning_rate": 3.0622600936279914e-06,
      "loss": 0.5461,
      "step": 880
    },
    {
      "epoch": 0.6675466716952668,
      "grad_norm": 2.0687230926828835,
      "learning_rate": 3.0016902550745896e-06,
      "loss": 0.5315,
      "step": 885
    },
    {
      "epoch": 0.671318121817839,
      "grad_norm": 2.1182888988383146,
      "learning_rate": 2.941467427938981e-06,
      "loss": 0.5251,
      "step": 890
    },
    {
      "epoch": 0.6750895719404111,
      "grad_norm": 2.166050854772299,
      "learning_rate": 2.881602070063695e-06,
      "loss": 0.5313,
      "step": 895
    },
    {
      "epoch": 0.6788610220629833,
      "grad_norm": 2.119609605339128,
      "learning_rate": 2.822104577215831e-06,
      "loss": 0.5298,
      "step": 900
    },
    {
      "epoch": 0.6826324721855553,
      "grad_norm": 2.3123977751286895,
      "learning_rate": 2.762985281281828e-06,
      "loss": 0.5162,
      "step": 905
    },
    {
      "epoch": 0.6864039223081275,
      "grad_norm": 2.147092195126045,
      "learning_rate": 2.7042544484732958e-06,
      "loss": 0.5225,
      "step": 910
    },
    {
      "epoch": 0.6901753724306996,
      "grad_norm": 2.1358012203129064,
      "learning_rate": 2.6459222775442582e-06,
      "loss": 0.5083,
      "step": 915
    },
    {
      "epoch": 0.6939468225532718,
      "grad_norm": 2.086154110222615,
      "learning_rate": 2.587998898020131e-06,
      "loss": 0.5163,
      "step": 920
    },
    {
      "epoch": 0.6977182726758439,
      "grad_norm": 2.1069385491797905,
      "learning_rate": 2.530494368438683e-06,
      "loss": 0.5131,
      "step": 925
    },
    {
      "epoch": 0.701489722798416,
      "grad_norm": 2.0817537626792757,
      "learning_rate": 2.473418674603363e-06,
      "loss": 0.5121,
      "step": 930
    },
    {
      "epoch": 0.7052611729209881,
      "grad_norm": 2.177796705379638,
      "learning_rate": 2.4167817278492337e-06,
      "loss": 0.5085,
      "step": 935
    },
    {
      "epoch": 0.7090326230435603,
      "grad_norm": 2.0422929484726238,
      "learning_rate": 2.3605933633218396e-06,
      "loss": 0.5126,
      "step": 940
    },
    {
      "epoch": 0.7128040731661324,
      "grad_norm": 2.0950516494712907,
      "learning_rate": 2.304863338269326e-06,
      "loss": 0.5019,
      "step": 945
    },
    {
      "epoch": 0.7165755232887046,
      "grad_norm": 2.139347400411628,
      "learning_rate": 2.249601330348044e-06,
      "loss": 0.5181,
      "step": 950
    },
    {
      "epoch": 0.7203469734112766,
      "grad_norm": 2.0787373792617103,
      "learning_rate": 2.194816935942024e-06,
      "loss": 0.5095,
      "step": 955
    },
    {
      "epoch": 0.7241184235338488,
      "grad_norm": 2.0934162351579007,
      "learning_rate": 2.140519668496529e-06,
      "loss": 0.5155,
      "step": 960
    },
    {
      "epoch": 0.7278898736564209,
      "grad_norm": 2.118132732572834,
      "learning_rate": 2.086718956866024e-06,
      "loss": 0.4989,
      "step": 965
    },
    {
      "epoch": 0.7316613237789931,
      "grad_norm": 2.015918946806562,
      "learning_rate": 2.033424143676825e-06,
      "loss": 0.4949,
      "step": 970
    },
    {
      "epoch": 0.7354327739015651,
      "grad_norm": 2.270252257697222,
      "learning_rate": 1.9806444837047478e-06,
      "loss": 0.5106,
      "step": 975
    },
    {
      "epoch": 0.7392042240241373,
      "grad_norm": 2.124764638447049,
      "learning_rate": 1.928389142267969e-06,
      "loss": 0.5041,
      "step": 980
    },
    {
      "epoch": 0.7429756741467094,
      "grad_norm": 2.047944456908165,
      "learning_rate": 1.8766671936354647e-06,
      "loss": 0.5092,
      "step": 985
    },
    {
      "epoch": 0.7467471242692816,
      "grad_norm": 2.0523494780642264,
      "learning_rate": 1.8254876194512344e-06,
      "loss": 0.4972,
      "step": 990
    },
    {
      "epoch": 0.7505185743918537,
      "grad_norm": 2.070102714868304,
      "learning_rate": 1.7748593071746089e-06,
      "loss": 0.516,
      "step": 995
    },
    {
      "epoch": 0.7542900245144258,
      "grad_norm": 2.1874025056879978,
      "learning_rate": 1.7247910485369417e-06,
      "loss": 0.504,
      "step": 1000
    },
    {
      "epoch": 0.7580614746369979,
      "grad_norm": 2.1407434998612547,
      "learning_rate": 1.6752915380148772e-06,
      "loss": 0.4958,
      "step": 1005
    },
    {
      "epoch": 0.7618329247595701,
      "grad_norm": 2.090141627816375,
      "learning_rate": 1.626369371320553e-06,
      "loss": 0.4954,
      "step": 1010
    },
    {
      "epoch": 0.7656043748821422,
      "grad_norm": 2.0433900279296147,
      "learning_rate": 1.578033043908928e-06,
      "loss": 0.4994,
      "step": 1015
    },
    {
      "epoch": 0.7693758250047144,
      "grad_norm": 2.3054787165813257,
      "learning_rate": 1.5302909495025164e-06,
      "loss": 0.4827,
      "step": 1020
    },
    {
      "epoch": 0.7731472751272864,
      "grad_norm": 2.07475519454638,
      "learning_rate": 1.4831513786338126e-06,
      "loss": 0.4983,
      "step": 1025
    },
    {
      "epoch": 0.7769187252498586,
      "grad_norm": 2.1363918208165686,
      "learning_rate": 1.4366225172056085e-06,
      "loss": 0.4939,
      "step": 1030
    },
    {
      "epoch": 0.7806901753724307,
      "grad_norm": 2.3053265439992034,
      "learning_rate": 1.390712445069496e-06,
      "loss": 0.4903,
      "step": 1035
    },
    {
      "epoch": 0.7844616254950029,
      "grad_norm": 2.0769867503422765,
      "learning_rate": 1.3454291346227787e-06,
      "loss": 0.4974,
      "step": 1040
    },
    {
      "epoch": 0.7882330756175749,
      "grad_norm": 2.1092355082866368,
      "learning_rate": 1.3007804494240478e-06,
      "loss": 0.4926,
      "step": 1045
    },
    {
      "epoch": 0.7920045257401471,
      "grad_norm": 2.0675161012597387,
      "learning_rate": 1.2567741428276513e-06,
      "loss": 0.4876,
      "step": 1050
    },
    {
      "epoch": 0.7957759758627192,
      "grad_norm": 2.1019083361876274,
      "learning_rate": 1.213417856637311e-06,
      "loss": 0.4805,
      "step": 1055
    },
    {
      "epoch": 0.7995474259852914,
      "grad_norm": 2.044608822527271,
      "learning_rate": 1.1707191197791012e-06,
      "loss": 0.488,
      "step": 1060
    },
    {
      "epoch": 0.8033188761078635,
      "grad_norm": 2.0934676621380035,
      "learning_rate": 1.128685346994028e-06,
      "loss": 0.4805,
      "step": 1065
    },
    {
      "epoch": 0.8070903262304356,
      "grad_norm": 2.168897054466299,
      "learning_rate": 1.0873238375504536e-06,
      "loss": 0.4966,
      "step": 1070
    },
    {
      "epoch": 0.8108617763530077,
      "grad_norm": 2.0831295415470694,
      "learning_rate": 1.046641773976544e-06,
      "loss": 0.4888,
      "step": 1075
    },
    {
      "epoch": 0.8146332264755799,
      "grad_norm": 2.0591460741221517,
      "learning_rate": 1.0066462208130174e-06,
      "loss": 0.4808,
      "step": 1080
    },
    {
      "epoch": 0.818404676598152,
      "grad_norm": 2.1413757269828677,
      "learning_rate": 9.673441233863661e-07,
      "loss": 0.4846,
      "step": 1085
    },
    {
      "epoch": 0.8221761267207242,
      "grad_norm": 2.0597898524385085,
      "learning_rate": 9.28742306602779e-07,
      "loss": 0.4801,
      "step": 1090
    },
    {
      "epoch": 0.8259475768432962,
      "grad_norm": 2.0619533474846072,
      "learning_rate": 8.908474737629897e-07,
      "loss": 0.4818,
      "step": 1095
    },
    {
      "epoch": 0.8297190269658684,
      "grad_norm": 2.0785020873929563,
      "learning_rate": 8.536662053982181e-07,
      "loss": 0.4796,
      "step": 1100
    },
    {
      "epoch": 0.8334904770884405,
      "grad_norm": 2.0473348850093886,
      "learning_rate": 8.172049581274571e-07,
      "loss": 0.4818,
      "step": 1105
    },
    {
      "epoch": 0.8372619272110127,
      "grad_norm": 2.0982171571132793,
      "learning_rate": 7.814700635362598e-07,
      "loss": 0.473,
      "step": 1110
    },
    {
      "epoch": 0.8410333773335847,
      "grad_norm": 2.1276785014387114,
      "learning_rate": 7.464677270772441e-07,
      "loss": 0.4779,
      "step": 1115
    },
    {
      "epoch": 0.8448048274561569,
      "grad_norm": 2.1127490084192146,
      "learning_rate": 7.122040269924985e-07,
      "loss": 0.4741,
      "step": 1120
    },
    {
      "epoch": 0.848576277578729,
      "grad_norm": 2.165957082369384,
      "learning_rate": 6.786849132580841e-07,
      "loss": 0.4797,
      "step": 1125
    },
    {
      "epoch": 0.8523477277013012,
      "grad_norm": 2.0518592705151115,
      "learning_rate": 6.459162065508067e-07,
      "loss": 0.478,
      "step": 1130
    },
    {
      "epoch": 0.8561191778238733,
      "grad_norm": 2.166898922792351,
      "learning_rate": 6.139035972374374e-07,
      "loss": 0.4817,
      "step": 1135
    },
    {
      "epoch": 0.8598906279464454,
      "grad_norm": 2.0882349445622808,
      "learning_rate": 5.826526443865732e-07,
      "loss": 0.4774,
      "step": 1140
    },
    {
      "epoch": 0.8636620780690175,
      "grad_norm": 2.0860313130166364,
      "learning_rate": 5.521687748032805e-07,
      "loss": 0.4808,
      "step": 1145
    },
    {
      "epoch": 0.8674335281915897,
      "grad_norm": 2.178027486209201,
      "learning_rate": 5.224572820867324e-07,
      "loss": 0.4818,
      "step": 1150
    },
    {
      "epoch": 0.8712049783141618,
      "grad_norm": 2.147704424651778,
      "learning_rate": 4.935233257109456e-07,
      "loss": 0.4768,
      "step": 1155
    },
    {
      "epoch": 0.874976428436734,
      "grad_norm": 2.070907197952466,
      "learning_rate": 4.653719301288362e-07,
      "loss": 0.4876,
      "step": 1160
    },
    {
      "epoch": 0.878747878559306,
      "grad_norm": 2.052204629034235,
      "learning_rate": 4.380079838997087e-07,
      "loss": 0.4746,
      "step": 1165
    },
    {
      "epoch": 0.8825193286818782,
      "grad_norm": 2.0244144715714176,
      "learning_rate": 4.114362388403409e-07,
      "loss": 0.4804,
      "step": 1170
    },
    {
      "epoch": 0.8862907788044503,
      "grad_norm": 1.997146927651021,
      "learning_rate": 3.8566130919982567e-07,
      "loss": 0.4711,
      "step": 1175
    },
    {
      "epoch": 0.8900622289270225,
      "grad_norm": 2.0695766404157574,
      "learning_rate": 3.606876708582918e-07,
      "loss": 0.4735,
      "step": 1180
    },
    {
      "epoch": 0.8938336790495945,
      "grad_norm": 2.088709302750632,
      "learning_rate": 3.3651966054965734e-07,
      "loss": 0.4811,
      "step": 1185
    },
    {
      "epoch": 0.8976051291721667,
      "grad_norm": 2.101334710434882,
      "learning_rate": 3.1316147510854434e-07,
      "loss": 0.4696,
      "step": 1190
    },
    {
      "epoch": 0.9013765792947388,
      "grad_norm": 2.085312151221947,
      "learning_rate": 2.9061717074148623e-07,
      "loss": 0.4763,
      "step": 1195
    },
    {
      "epoch": 0.905148029417311,
      "grad_norm": 2.139678602127972,
      "learning_rate": 2.688906623225629e-07,
      "loss": 0.4777,
      "step": 1200
    },
    {
      "epoch": 0.9089194795398831,
      "grad_norm": 2.1249347744645086,
      "learning_rate": 2.479857227135685e-07,
      "loss": 0.4709,
      "step": 1205
    },
    {
      "epoch": 0.9126909296624552,
      "grad_norm": 2.108323773956457,
      "learning_rate": 2.2790598210884695e-07,
      "loss": 0.476,
      "step": 1210
    },
    {
      "epoch": 0.9164623797850273,
      "grad_norm": 2.0522415633167,
      "learning_rate": 2.0865492740489822e-07,
      "loss": 0.4663,
      "step": 1215
    },
    {
      "epoch": 0.9202338299075995,
      "grad_norm": 2.015502653559595,
      "learning_rate": 1.9023590159487237e-07,
      "loss": 0.4678,
      "step": 1220
    },
    {
      "epoch": 0.9240052800301716,
      "grad_norm": 2.1033783508867647,
      "learning_rate": 1.7265210318804683e-07,
      "loss": 0.4735,
      "step": 1225
    },
    {
      "epoch": 0.9277767301527438,
      "grad_norm": 2.0849738071956017,
      "learning_rate": 1.5590658565440063e-07,
      "loss": 0.4814,
      "step": 1230
    },
    {
      "epoch": 0.9315481802753158,
      "grad_norm": 2.1194015599967018,
      "learning_rate": 1.4000225689437263e-07,
      "loss": 0.4767,
      "step": 1235
    },
    {
      "epoch": 0.935319630397888,
      "grad_norm": 2.155785238683729,
      "learning_rate": 1.2494187873389186e-07,
      "loss": 0.4702,
      "step": 1240
    },
    {
      "epoch": 0.9390910805204601,
      "grad_norm": 2.1492733094402428,
      "learning_rate": 1.107280664447874e-07,
      "loss": 0.4678,
      "step": 1245
    },
    {
      "epoch": 0.9428625306430323,
      "grad_norm": 2.0766452818387973,
      "learning_rate": 9.73632882906328e-08,
      "loss": 0.4716,
      "step": 1250
    },
    {
      "epoch": 0.9466339807656043,
      "grad_norm": 2.0669365458508513,
      "learning_rate": 8.484986509813165e-08,
      "loss": 0.4651,
      "step": 1255
    },
    {
      "epoch": 0.9504054308881765,
      "grad_norm": 2.1012750712142587,
      "learning_rate": 7.318996985409609e-08,
      "loss": 0.4744,
      "step": 1260
    },
    {
      "epoch": 0.9541768810107486,
      "grad_norm": 2.0562321019527188,
      "learning_rate": 6.238562732810427e-08,
      "loss": 0.4626,
      "step": 1265
    },
    {
      "epoch": 0.9579483311333208,
      "grad_norm": 2.2665109481569305,
      "learning_rate": 5.243871372089482e-08,
      "loss": 0.4756,
      "step": 1270
    },
    {
      "epoch": 0.9617197812558929,
      "grad_norm": 2.01339402739892,
      "learning_rate": 4.335095633855868e-08,
      "loss": 0.4701,
      "step": 1275
    },
    {
      "epoch": 0.965491231378465,
      "grad_norm": 2.0355724329334057,
      "learning_rate": 3.512393329258734e-08,
      "loss": 0.4681,
      "step": 1280
    },
    {
      "epoch": 0.9692626815010371,
      "grad_norm": 2.0763696915453793,
      "learning_rate": 2.77590732258326e-08,
      "loss": 0.4853,
      "step": 1285
    },
    {
      "epoch": 0.9730341316236093,
      "grad_norm": 2.223079040055733,
      "learning_rate": 2.1257655064416684e-08,
      "loss": 0.4755,
      "step": 1290
    },
    {
      "epoch": 0.9768055817461814,
      "grad_norm": 2.0350565583163522,
      "learning_rate": 1.56208077956449e-08,
      "loss": 0.4644,
      "step": 1295
    },
    {
      "epoch": 0.9805770318687536,
      "grad_norm": 2.0017384058730805,
      "learning_rate": 1.0849510271953578e-08,
      "loss": 0.4785,
      "step": 1300
    },
    {
      "epoch": 0.9843484819913256,
      "grad_norm": 2.0937310226370687,
      "learning_rate": 6.944591040930481e-09,
      "loss": 0.4697,
      "step": 1305
    },
    {
      "epoch": 0.9881199321138978,
      "grad_norm": 2.0794638687177938,
      "learning_rate": 3.9067282014354504e-09,
      "loss": 0.4707,
      "step": 1310
    },
    {
      "epoch": 0.9918913822364699,
      "grad_norm": 2.0578690852807098,
      "learning_rate": 1.7364492858451542e-09,
      "loss": 0.468,
      "step": 1315
    },
    {
      "epoch": 0.9956628323590421,
      "grad_norm": 2.1418033123461093,
      "learning_rate": 4.341311684485927e-10,
      "loss": 0.4788,
      "step": 1320
    },
    {
      "epoch": 0.9994342824816141,
      "grad_norm": 2.041628364481552,
      "learning_rate": 0.0,
      "loss": 0.4741,
      "step": 1325
    },
    {
      "epoch": 0.9994342824816141,
      "eval_loss": 0.47078943252563477,
      "eval_runtime": 251.4407,
      "eval_samples_per_second": 3.094,
      "eval_steps_per_second": 0.776,
      "step": 1325
    },
    {
      "epoch": 0.9994342824816141,
      "step": 1325,
      "total_flos": 277375698862080.0,
      "train_loss": 0.6321603435840247,
      "train_runtime": 30135.1647,
      "train_samples_per_second": 1.408,
      "train_steps_per_second": 0.044
    }
  ],
  "logging_steps": 5,
  "max_steps": 1325,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 277375698862080.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}