{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 2.9881796836853027,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0205,
      "step": 10
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6945570707321167,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9355,
      "step": 20
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0729589462280273,
      "learning_rate": 1e-05,
      "loss": 0.8719,
      "step": 30
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8945010900497437,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.7976,
      "step": 40
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9784910678863525,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.8037,
      "step": 50
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9404758810997009,
      "learning_rate": 2e-05,
      "loss": 0.7779,
      "step": 60
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.9830419421195984,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 0.779,
      "step": 70
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.0543450117111206,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.7676,
      "step": 80
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.151178002357483,
      "learning_rate": 3e-05,
      "loss": 0.7662,
      "step": 90
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1893316507339478,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.7598,
      "step": 100
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.1506898403167725,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 0.7463,
      "step": 110
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2127164602279663,
      "learning_rate": 4e-05,
      "loss": 0.749,
      "step": 120
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.0937416553497314,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 0.7379,
      "step": 130
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.172784686088562,
      "learning_rate": 4.666666666666667e-05,
      "loss": 0.7427,
      "step": 140
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1628010272979736,
      "learning_rate": 5e-05,
      "loss": 0.7444,
      "step": 150
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.4532078504562378,
      "learning_rate": 4.999848114735858e-05,
      "loss": 0.7257,
      "step": 160
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1651058197021484,
      "learning_rate": 4.999392477398737e-05,
      "loss": 0.7348,
      "step": 170
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.134162425994873,
      "learning_rate": 4.9986331433523156e-05,
      "loss": 0.7136,
      "step": 180
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.1000313758850098,
      "learning_rate": 4.997570204861915e-05,
      "loss": 0.7167,
      "step": 190
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.0679141283035278,
      "learning_rate": 4.996203791083291e-05,
      "loss": 0.7172,
      "step": 200
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.1267646551132202,
      "learning_rate": 4.994534068046937e-05,
      "loss": 0.7276,
      "step": 210
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.1281580924987793,
      "learning_rate": 4.992561238637912e-05,
      "loss": 0.7187,
      "step": 220
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.9922609329223633,
      "learning_rate": 4.9902855425711905e-05,
      "loss": 0.7085,
      "step": 230
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.9926589131355286,
      "learning_rate": 4.9877072563625285e-05,
      "loss": 0.7185,
      "step": 240
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0269533395767212,
      "learning_rate": 4.984826693294874e-05,
      "loss": 0.7208,
      "step": 250
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.9333322644233704,
      "learning_rate": 4.981644203380291e-05,
      "loss": 0.7052,
      "step": 260
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.9695472717285156,
      "learning_rate": 4.978160173317438e-05,
      "loss": 0.695,
      "step": 270
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.0205656290054321,
      "learning_rate": 4.974375026444575e-05,
      "loss": 0.7029,
      "step": 280
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.8118201494216919,
      "learning_rate": 4.970289222688129e-05,
      "loss": 0.6981,
      "step": 290
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.9229862093925476,
      "learning_rate": 4.965903258506806e-05,
      "loss": 0.7108,
      "step": 300
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9452656507492065,
      "learning_rate": 4.961217666831268e-05,
      "loss": 0.6916,
      "step": 310
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.9605855345726013,
      "learning_rate": 4.956233016999379e-05,
      "loss": 0.696,
      "step": 320
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.8793039917945862,
      "learning_rate": 4.9509499146870236e-05,
      "loss": 0.6876,
      "step": 330
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.8802406787872314,
      "learning_rate": 4.9453690018345144e-05,
      "loss": 0.6996,
      "step": 340
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.0251939296722412,
      "learning_rate": 4.9394909565685894e-05,
      "loss": 0.7063,
      "step": 350
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.8812974691390991,
      "learning_rate": 4.933316493120015e-05,
      "loss": 0.6948,
      "step": 360
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7881792783737183,
      "learning_rate": 4.9268463617368e-05,
      "loss": 0.6951,
      "step": 370
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.8760414123535156,
      "learning_rate": 4.9200813485930375e-05,
      "loss": 0.6906,
      "step": 380
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.8838094472885132,
      "learning_rate": 4.913022275693372e-05,
      "loss": 0.6772,
      "step": 390
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.0329331159591675,
      "learning_rate": 4.905670000773126e-05,
      "loss": 0.6986,
      "step": 400
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.7983132004737854,
      "learning_rate": 4.8980254171940746e-05,
      "loss": 0.6811,
      "step": 410
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9415234327316284,
      "learning_rate": 4.8900894538358944e-05,
      "loss": 0.6697,
      "step": 420
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.76316237449646,
      "learning_rate": 4.881863074983298e-05,
      "loss": 0.6851,
      "step": 430
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.8890565633773804,
      "learning_rate": 4.8733472802088654e-05,
      "loss": 0.6819,
      "step": 440
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.7978311777114868,
      "learning_rate": 4.864543104251587e-05,
      "loss": 0.6852,
      "step": 450
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.8316269516944885,
      "learning_rate": 4.855451616891136e-05,
      "loss": 0.6718,
      "step": 460
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.877873957157135,
      "learning_rate": 4.8460739228178806e-05,
      "loss": 0.6707,
      "step": 470
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.0596739053726196,
      "learning_rate": 4.8364111614986527e-05,
      "loss": 0.6964,
      "step": 480
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.8643849492073059,
      "learning_rate": 4.8264645070382964e-05,
      "loss": 0.6763,
      "step": 490
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7877243161201477,
      "learning_rate": 4.8162351680370044e-05,
      "loss": 0.6874,
      "step": 500
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9450274109840393,
      "learning_rate": 4.805724387443462e-05,
      "loss": 0.6818,
      "step": 510
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.7456744313240051,
      "learning_rate": 4.7949334424038176e-05,
      "loss": 0.665,
      "step": 520
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8591768741607666,
      "learning_rate": 4.783863644106502e-05,
      "loss": 0.6675,
      "step": 530
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.7636451125144958,
      "learning_rate": 4.7725163376229064e-05,
      "loss": 0.6633,
      "step": 540
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.8254076242446899,
      "learning_rate": 4.760892901743944e-05,
      "loss": 0.667,
      "step": 550
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.8828392624855042,
      "learning_rate": 4.7489947488125175e-05,
      "loss": 0.6842,
      "step": 560
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.8078437447547913,
      "learning_rate": 4.736823324551909e-05,
      "loss": 0.665,
      "step": 570
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.8083636164665222,
      "learning_rate": 4.7243801078901084e-05,
      "loss": 0.6577,
      "step": 580
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9294388890266418,
      "learning_rate": 4.711666610780115e-05,
      "loss": 0.668,
      "step": 590
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8191084265708923,
      "learning_rate": 4.698684378016222e-05,
      "loss": 0.6571,
      "step": 600
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7333921194076538,
      "learning_rate": 4.685434987046314e-05,
      "loss": 0.6671,
      "step": 610
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.7950130701065063,
      "learning_rate": 4.671920047780186e-05,
      "loss": 0.6549,
      "step": 620
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8252810835838318,
      "learning_rate": 4.6581412023939354e-05,
      "loss": 0.6517,
      "step": 630
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7834108471870422,
      "learning_rate": 4.644100125130418e-05,
      "loss": 0.6707,
      "step": 640
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.7022207379341125,
      "learning_rate": 4.629798522095818e-05,
      "loss": 0.6698,
      "step": 650
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8147993087768555,
      "learning_rate": 4.6152381310523387e-05,
      "loss": 0.651,
      "step": 660
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9004583358764648,
      "learning_rate": 4.600420721207053e-05,
      "loss": 0.6694,
      "step": 670
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.7641873955726624,
      "learning_rate": 4.585348092996925e-05,
      "loss": 0.655,
      "step": 680
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.7779438495635986,
      "learning_rate": 4.5700220778700504e-05,
      "loss": 0.6514,
      "step": 690
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.7963545322418213,
      "learning_rate": 4.554444538063113e-05,
      "loss": 0.635,
      "step": 700
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.7627516388893127,
      "learning_rate": 4.538617366375112e-05,
      "loss": 0.6472,
      "step": 710
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.6603427529335022,
      "learning_rate": 4.522542485937369e-05,
      "loss": 0.6416,
      "step": 720
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.7783029079437256,
      "learning_rate": 4.5062218499798526e-05,
      "loss": 0.644,
      "step": 730
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.7225221395492554,
      "learning_rate": 4.4896574415938465e-05,
      "loss": 0.644,
      "step": 740
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8217127919197083,
      "learning_rate": 4.4728512734909844e-05,
      "loss": 0.647,
      "step": 750
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9464953541755676,
      "learning_rate": 4.455805387758691e-05,
      "loss": 0.6629,
      "step": 760
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.7954788208007812,
      "learning_rate": 4.438521855612054e-05,
      "loss": 0.6603,
      "step": 770
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.7955674529075623,
      "learning_rate": 4.421002777142148e-05,
      "loss": 0.6616,
      "step": 780
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.7647875547409058,
      "learning_rate": 4.4032502810608614e-05,
      "loss": 0.6463,
      "step": 790
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.7553123235702515,
      "learning_rate": 4.385266524442241e-05,
      "loss": 0.6453,
      "step": 800
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.6640244722366333,
      "learning_rate": 4.367053692460385e-05,
      "loss": 0.6269,
      "step": 810
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.7737441062927246,
      "learning_rate": 4.3486139981239304e-05,
      "loss": 0.6546,
      "step": 820
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.6590189337730408,
      "learning_rate": 4.3299496820071546e-05,
      "loss": 0.6414,
      "step": 830
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.7308911681175232,
      "learning_rate": 4.311063011977723e-05,
      "loss": 0.6435,
      "step": 840
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.7682307362556458,
      "learning_rate": 4.2919562829211283e-05,
      "loss": 0.6416,
      "step": 850
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.7915315628051758,
      "learning_rate": 4.2726318164618435e-05,
      "loss": 0.6504,
      "step": 860
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.7202879190444946,
      "learning_rate": 4.2530919606812216e-05,
      "loss": 0.6491,
      "step": 870
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.6950281858444214,
      "learning_rate": 4.233339089832189e-05,
      "loss": 0.6422,
      "step": 880
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.715614914894104,
      "learning_rate": 4.21337560405075e-05,
      "loss": 0.6461,
      "step": 890
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.6588935852050781,
      "learning_rate": 4.193203929064353e-05,
      "loss": 0.6248,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.6951584815979004,
      "learning_rate": 4.172826515897146e-05,
      "loss": 0.6415,
      "step": 910
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.7861813306808472,
      "learning_rate": 4.152245840572153e-05,
      "loss": 0.6359,
      "step": 920
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.7668294310569763,
      "learning_rate": 4.131464403810422e-05,
      "loss": 0.6557,
      "step": 930
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.7085067629814148,
      "learning_rate": 4.110484730727161e-05,
      "loss": 0.6417,
      "step": 940
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.7475394010543823,
      "learning_rate": 4.089309370524921e-05,
      "loss": 0.624,
      "step": 950
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.7203657627105713,
      "learning_rate": 4.067940896183843e-05,
      "loss": 0.628,
      "step": 960
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.7134045958518982,
      "learning_rate": 4.046381904149024e-05,
      "loss": 0.6346,
      "step": 970
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.6537560820579529,
      "learning_rate": 4.024635014015023e-05,
      "loss": 0.6259,
      "step": 980
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.7190784215927124,
      "learning_rate": 4.002702868207563e-05,
      "loss": 0.6373,
      "step": 990
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6727094650268555,
      "learning_rate": 3.9805881316624506e-05,
      "loss": 0.6158,
      "step": 1000
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.7174749970436096,
      "learning_rate": 3.9582934915017665e-05,
      "loss": 0.6011,
      "step": 1010
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.6894753575325012,
      "learning_rate": 3.935821656707359e-05,
      "loss": 0.5974,
      "step": 1020
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.6777900457382202,
      "learning_rate": 3.91317535779168e-05,
      "loss": 0.6027,
      "step": 1030
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.680266797542572,
      "learning_rate": 3.890357346466001e-05,
      "loss": 0.5857,
      "step": 1040
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.7262083292007446,
      "learning_rate": 3.867370395306068e-05,
      "loss": 0.5985,
      "step": 1050
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.6810183525085449,
      "learning_rate": 3.844217297415196e-05,
      "loss": 0.5874,
      "step": 1060
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.743713915348053,
      "learning_rate": 3.8209008660848974e-05,
      "loss": 0.5842,
      "step": 1070
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.7603400349617004,
      "learning_rate": 3.797423934453038e-05,
      "loss": 0.596,
      "step": 1080
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.6763606071472168,
      "learning_rate": 3.773789355159587e-05,
      "loss": 0.6043,
      "step": 1090
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.7692492008209229,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.599,
      "step": 1100
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.750664472579956,
      "learning_rate": 3.726058759576271e-05,
      "loss": 0.5878,
      "step": 1110
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.6852485537528992,
      "learning_rate": 3.7019685429456986e-05,
      "loss": 0.5905,
      "step": 1120
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.7305880188941956,
      "learning_rate": 3.6777322772674186e-05,
      "loss": 0.5977,
      "step": 1130
    },
    {
      "epoch": 1.1400000000000001,
      "grad_norm": 0.6685648560523987,
      "learning_rate": 3.65335290744672e-05,
      "loss": 0.5987,
      "step": 1140
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.730179488658905,
      "learning_rate": 3.628833395777224e-05,
      "loss": 0.5957,
      "step": 1150
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.6852392554283142,
      "learning_rate": 3.604176721580935e-05,
      "loss": 0.5961,
      "step": 1160
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.6334732174873352,
      "learning_rate": 3.579385880846232e-05,
      "loss": 0.5793,
      "step": 1170
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.6806017756462097,
      "learning_rate": 3.5544638858638304e-05,
      "loss": 0.5969,
      "step": 1180
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.6200957894325256,
      "learning_rate": 3.5294137648607625e-05,
      "loss": 0.5973,
      "step": 1190
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.6411502361297607,
      "learning_rate": 3.504238561632424e-05,
      "loss": 0.5812,
      "step": 1200
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.622643232345581,
      "learning_rate": 3.478941335172729e-05,
      "loss": 0.5891,
      "step": 1210
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.6710910797119141,
      "learning_rate": 3.453525159302415e-05,
      "loss": 0.6081,
      "step": 1220
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.6420513987541199,
      "learning_rate": 3.427993122295552e-05,
      "loss": 0.6082,
      "step": 1230
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.6355459690093994,
      "learning_rate": 3.4023483265042874e-05,
      "loss": 0.5871,
      "step": 1240
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.7996178269386292,
      "learning_rate": 3.376593887981887e-05,
      "loss": 0.6023,
      "step": 1250
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.7436390519142151,
      "learning_rate": 3.350732936104108e-05,
      "loss": 0.5881,
      "step": 1260
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.6650518774986267,
      "learning_rate": 3.3247686131889574e-05,
      "loss": 0.5856,
      "step": 1270
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.6648463606834412,
      "learning_rate": 3.29870407411487e-05,
      "loss": 0.5995,
      "step": 1280
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.6547637581825256,
      "learning_rate": 3.272542485937369e-05,
      "loss": 0.5896,
      "step": 1290
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.6670102477073669,
      "learning_rate": 3.246287027504237e-05,
      "loss": 0.5846,
      "step": 1300
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.6504084467887878,
      "learning_rate": 3.2199408890692655e-05,
      "loss": 0.5891,
      "step": 1310
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.7213049530982971,
      "learning_rate": 3.1935072719046115e-05,
      "loss": 0.5899,
      "step": 1320
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.6338471174240112,
      "learning_rate": 3.1669893879118156e-05,
      "loss": 0.587,
      "step": 1330
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.6446500420570374,
      "learning_rate": 3.140390459231528e-05,
      "loss": 0.5862,
      "step": 1340
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.626853346824646,
      "learning_rate": 3.1137137178519985e-05,
      "loss": 0.5982,
      "step": 1350
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 0.6430269479751587,
      "learning_rate": 3.086962405216353e-05,
      "loss": 0.592,
      "step": 1360
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.590183436870575,
      "learning_rate": 3.06013977182874e-05,
      "loss": 0.5848,
      "step": 1370
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.6320422291755676,
      "learning_rate": 3.0332490768593675e-05,
      "loss": 0.5794,
      "step": 1380
    },
    {
      "epoch": 1.3900000000000001,
      "grad_norm": 0.5953449010848999,
      "learning_rate": 3.0062935877484804e-05,
      "loss": 0.6042,
      "step": 1390
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.5875259041786194,
      "learning_rate": 2.9792765798093465e-05,
      "loss": 0.6037,
      "step": 1400
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.6373878717422485,
      "learning_rate": 2.952201335830275e-05,
      "loss": 0.5921,
      "step": 1410
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.7542224526405334,
      "learning_rate": 2.925071145675733e-05,
      "loss": 0.5948,
      "step": 1420
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.6458872556686401,
      "learning_rate": 2.8978893058865987e-05,
      "loss": 0.5762,
      "step": 1430
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.741593062877655,
      "learning_rate": 2.870659119279605e-05,
      "loss": 0.6207,
      "step": 1440
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.6762745976448059,
      "learning_rate": 2.8433838945460205e-05,
      "loss": 0.5978,
      "step": 1450
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.6076602339744568,
      "learning_rate": 2.8160669458496158e-05,
      "loss": 0.5937,
      "step": 1460
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.556096613407135,
      "learning_rate": 2.788711592423966e-05,
      "loss": 0.589,
      "step": 1470
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.710705578327179,
      "learning_rate": 2.761321158169134e-05,
      "loss": 0.5781,
      "step": 1480
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.6006096005439758,
      "learning_rate": 2.7338989712477945e-05,
      "loss": 0.5905,
      "step": 1490
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.6222957372665405,
      "learning_rate": 2.7064483636808313e-05,
      "loss": 0.5743,
      "step": 1500
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.6598724722862244,
      "learning_rate": 2.678972670942468e-05,
      "loss": 0.586,
      "step": 1510
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.6561234593391418,
      "learning_rate": 2.6514752315549847e-05,
      "loss": 0.5802,
      "step": 1520
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.591433048248291,
      "learning_rate": 2.623959386683056e-05,
      "loss": 0.5828,
      "step": 1530
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.6482975482940674,
      "learning_rate": 2.5964284797277762e-05,
      "loss": 0.5961,
      "step": 1540
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.6356632113456726,
      "learning_rate": 2.5688858559204053e-05,
      "loss": 0.5887,
      "step": 1550
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.6521331071853638,
      "learning_rate": 2.5413348619158967e-05,
      "loss": 0.5943,
      "step": 1560
    },
    {
      "epoch": 1.5699999999999998,
      "grad_norm": 0.589897096157074,
      "learning_rate": 2.5137788453862515e-05,
      "loss": 0.5925,
      "step": 1570
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.6734787225723267,
      "learning_rate": 2.486221154613749e-05,
      "loss": 0.5725,
      "step": 1580
    },
    {
      "epoch": 1.5899999999999999,
      "grad_norm": 0.613000214099884,
      "learning_rate": 2.458665138084104e-05,
      "loss": 0.5832,
      "step": 1590
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.6318116188049316,
      "learning_rate": 2.4311141440795953e-05,
      "loss": 0.5779,
      "step": 1600
    },
    {
      "epoch": 1.6099999999999999,
      "grad_norm": 0.6127640604972839,
      "learning_rate": 2.4035715202722237e-05,
      "loss": 0.575,
      "step": 1610
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.6314959526062012,
      "learning_rate": 2.3760406133169443e-05,
      "loss": 0.5818,
      "step": 1620
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.6150800585746765,
      "learning_rate": 2.3485247684450166e-05,
      "loss": 0.5743,
      "step": 1630
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 0.5990105867385864,
      "learning_rate": 2.3210273290575333e-05,
      "loss": 0.5764,
      "step": 1640
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.6026667356491089,
      "learning_rate": 2.2935516363191693e-05,
      "loss": 0.5795,
      "step": 1650
    },
    {
      "epoch": 1.6600000000000001,
      "grad_norm": 0.614840567111969,
      "learning_rate": 2.2661010287522057e-05,
      "loss": 0.5821,
      "step": 1660
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.5916484594345093,
      "learning_rate": 2.238678841830867e-05,
      "loss": 0.5819,
      "step": 1670
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 0.5516738891601562,
      "learning_rate": 2.2112884075760347e-05,
      "loss": 0.5835,
      "step": 1680
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.582239031791687,
      "learning_rate": 2.1839330541503845e-05,
      "loss": 0.5751,
      "step": 1690
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.5710985660552979,
      "learning_rate": 2.1566161054539798e-05,
      "loss": 0.5777,
      "step": 1700
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.5882836580276489,
      "learning_rate": 2.1293408807203947e-05,
      "loss": 0.5787,
      "step": 1710
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.5639011859893799,
      "learning_rate": 2.1021106941134012e-05,
      "loss": 0.585,
      "step": 1720
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.6191072463989258,
      "learning_rate": 2.074928854324268e-05,
      "loss": 0.5872,
      "step": 1730
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.5924238562583923,
      "learning_rate": 2.047798664169726e-05,
      "loss": 0.5823,
      "step": 1740
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.537702202796936,
      "learning_rate": 2.0207234201906547e-05,
      "loss": 0.5744,
      "step": 1750
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.5480232238769531,
      "learning_rate": 1.9937064122515202e-05,
      "loss": 0.5909,
      "step": 1760
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.5865290760993958,
      "learning_rate": 1.9667509231406334e-05,
      "loss": 0.5691,
      "step": 1770
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.5969544649124146,
      "learning_rate": 1.9398602281712604e-05,
      "loss": 0.5763,
      "step": 1780
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.5609118342399597,
      "learning_rate": 1.913037594783648e-05,
      "loss": 0.5714,
      "step": 1790
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.5902943015098572,
      "learning_rate": 1.8862862821480025e-05,
      "loss": 0.5676,
      "step": 1800
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.6172893047332764,
      "learning_rate": 1.859609540768471e-05,
      "loss": 0.568,
      "step": 1810
    },
    {
      "epoch": 1.8199999999999998,
      "grad_norm": 0.6336326003074646,
      "learning_rate": 1.8330106120881846e-05,
      "loss": 0.582,
      "step": 1820
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.589616060256958,
      "learning_rate": 1.806492728095389e-05,
      "loss": 0.581,
      "step": 1830
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 0.5233524441719055,
      "learning_rate": 1.780059110930735e-05,
      "loss": 0.578,
      "step": 1840
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.5605369210243225,
      "learning_rate": 1.7537129724957642e-05,
      "loss": 0.5696,
      "step": 1850
    },
    {
      "epoch": 1.8599999999999999,
      "grad_norm": 0.5463587641716003,
      "learning_rate": 1.7274575140626318e-05,
      "loss": 0.5717,
      "step": 1860
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.5655060410499573,
      "learning_rate": 1.70129592588513e-05,
      "loss": 0.5698,
      "step": 1870
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.5800077319145203,
      "learning_rate": 1.675231386811043e-05,
      "loss": 0.5844,
      "step": 1880
    },
    {
      "epoch": 1.8900000000000001,
      "grad_norm": 0.5689355134963989,
      "learning_rate": 1.6492670638958924e-05,
      "loss": 0.5847,
      "step": 1890
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.576605498790741,
      "learning_rate": 1.6234061120181142e-05,
      "loss": 0.5783,
      "step": 1900
    },
    {
      "epoch": 1.9100000000000001,
      "grad_norm": 0.5839110612869263,
      "learning_rate": 1.5976516734957138e-05,
      "loss": 0.5676,
      "step": 1910
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.5562628507614136,
      "learning_rate": 1.5720068777044476e-05,
      "loss": 0.5684,
      "step": 1920
    },
    {
      "epoch": 1.9300000000000002,
      "grad_norm": 0.5681043863296509,
      "learning_rate": 1.5464748406975847e-05,
      "loss": 0.5769,
      "step": 1930
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.5721532702445984,
      "learning_rate": 1.521058664827272e-05,
      "loss": 0.5737,
      "step": 1940
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.538774311542511,
      "learning_rate": 1.495761438367577e-05,
      "loss": 0.5585,
      "step": 1950
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.5676820874214172,
      "learning_rate": 1.4705862351392379e-05,
      "loss": 0.5831,
      "step": 1960
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.5609452128410339,
      "learning_rate": 1.44553611413617e-05,
      "loss": 0.5728,
      "step": 1970
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.5539233088493347,
      "learning_rate": 1.4206141191537682e-05,
      "loss": 0.5724,
      "step": 1980
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.5074451565742493,
      "learning_rate": 1.395823278419065e-05,
      "loss": 0.5662,
      "step": 1990
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5767993927001953,
      "learning_rate": 1.3711666042227772e-05,
      "loss": 0.5646,
      "step": 2000
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.6038601398468018,
      "learning_rate": 1.346647092553281e-05,
      "loss": 0.5408,
      "step": 2010
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.5604304075241089,
      "learning_rate": 1.322267722732582e-05,
      "loss": 0.5433,
      "step": 2020
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.5742123126983643,
      "learning_rate": 1.2980314570543006e-05,
      "loss": 0.5506,
      "step": 2030
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.6020936965942383,
      "learning_rate": 1.2739412404237306e-05,
      "loss": 0.5405,
      "step": 2040
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.5823059678077698,
      "learning_rate": 1.2500000000000006e-05,
      "loss": 0.5317,
      "step": 2050
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.5085962414741516,
      "learning_rate": 1.2262106448404132e-05,
      "loss": 0.5315,
      "step": 2060
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.5282750129699707,
      "learning_rate": 1.202576065546963e-05,
      "loss": 0.5395,
      "step": 2070
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.5278365612030029,
      "learning_rate": 1.1790991339151031e-05,
      "loss": 0.5417,
      "step": 2080
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.556912362575531,
      "learning_rate": 1.1557827025848047e-05,
      "loss": 0.5389,
      "step": 2090
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.586982250213623,
      "learning_rate": 1.1326296046939333e-05,
      "loss": 0.5385,
      "step": 2100
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.5237417221069336,
      "learning_rate": 1.1096426535339985e-05,
      "loss": 0.5379,
      "step": 2110
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.5813203454017639,
      "learning_rate": 1.0868246422083204e-05,
      "loss": 0.5411,
      "step": 2120
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.5908067226409912,
      "learning_rate": 1.064178343292641e-05,
      "loss": 0.5283,
      "step": 2130
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.6426777243614197,
      "learning_rate": 1.0417065084982346e-05,
      "loss": 0.5427,
      "step": 2140
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.5757883787155151,
      "learning_rate": 1.0194118683375503e-05,
      "loss": 0.5306,
      "step": 2150
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.584816575050354,
      "learning_rate": 9.972971317924374e-06,
      "loss": 0.5222,
      "step": 2160
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.6235145330429077,
      "learning_rate": 9.753649859849775e-06,
      "loss": 0.5349,
      "step": 2170
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.5281869173049927,
      "learning_rate": 9.536180958509768e-06,
      "loss": 0.5352,
      "step": 2180
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.5590287446975708,
      "learning_rate": 9.320591038161574e-06,
      "loss": 0.5348,
      "step": 2190
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.5477201342582703,
      "learning_rate": 9.106906294750805e-06,
      "loss": 0.5213,
      "step": 2200
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.5789000988006592,
      "learning_rate": 8.895152692728397e-06,
      "loss": 0.5331,
      "step": 2210
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.5713198781013489,
      "learning_rate": 8.685355961895784e-06,
      "loss": 0.5403,
      "step": 2220
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.5491828322410583,
      "learning_rate": 8.477541594278474e-06,
      "loss": 0.5371,
      "step": 2230
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.576400876045227,
      "learning_rate": 8.271734841028553e-06,
      "loss": 0.5314,
      "step": 2240
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.5434856414794922,
      "learning_rate": 8.067960709356478e-06,
      "loss": 0.547,
      "step": 2250
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.5376303791999817,
      "learning_rate": 7.866243959492509e-06,
      "loss": 0.5326,
      "step": 2260
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.5279624462127686,
      "learning_rate": 7.666609101678121e-06,
      "loss": 0.5297,
      "step": 2270
    },
    {
      "epoch": 2.2800000000000002,
      "grad_norm": 0.5436615943908691,
      "learning_rate": 7.469080393187786e-06,
      "loss": 0.537,
      "step": 2280
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.4894987642765045,
      "learning_rate": 7.273681835381569e-06,
      "loss": 0.541,
      "step": 2290
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.5661471486091614,
      "learning_rate": 7.080437170788723e-06,
      "loss": 0.534,
      "step": 2300
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.5674607753753662,
      "learning_rate": 6.889369880222776e-06,
      "loss": 0.5305,
      "step": 2310
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.5481733083724976,
      "learning_rate": 6.700503179928458e-06,
      "loss": 0.5337,
      "step": 2320
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.5626810789108276,
      "learning_rate": 6.513860018760698e-06,
      "loss": 0.5391,
      "step": 2330
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.545527458190918,
      "learning_rate": 6.329463075396161e-06,
      "loss": 0.5442,
      "step": 2340
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.5546277761459351,
      "learning_rate": 6.147334755577596e-06,
      "loss": 0.5446,
      "step": 2350
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.5283880829811096,
      "learning_rate": 5.967497189391386e-06,
      "loss": 0.5246,
      "step": 2360
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.5393268465995789,
      "learning_rate": 5.78997222857853e-06,
      "loss": 0.535,
      "step": 2370
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.5287243127822876,
      "learning_rate": 5.614781443879463e-06,
      "loss": 0.5291,
      "step": 2380
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.5424278378486633,
      "learning_rate": 5.441946122413086e-06,
      "loss": 0.5302,
      "step": 2390
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.48928436636924744,
      "learning_rate": 5.271487265090163e-06,
      "loss": 0.5377,
      "step": 2400
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.5223726034164429,
      "learning_rate": 5.103425584061538e-06,
      "loss": 0.5235,
      "step": 2410
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.5279435515403748,
      "learning_rate": 4.937781500201474e-06,
      "loss": 0.524,
      "step": 2420
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.5740906596183777,
      "learning_rate": 4.7745751406263165e-06,
      "loss": 0.5433,
      "step": 2430
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.5629609823226929,
      "learning_rate": 4.613826336248881e-06,
      "loss": 0.5366,
      "step": 2440
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.5071305632591248,
      "learning_rate": 4.4555546193688735e-06,
      "loss": 0.5327,
      "step": 2450
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.5691395998001099,
      "learning_rate": 4.299779221299499e-06,
      "loss": 0.538,
      "step": 2460
    },
    {
      "epoch": 2.4699999999999998,
      "grad_norm": 0.5553451180458069,
      "learning_rate": 4.146519070030757e-06,
      "loss": 0.5254,
      "step": 2470
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.5273837447166443,
      "learning_rate": 3.995792787929481e-06,
      "loss": 0.5287,
      "step": 2480
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.5408598184585571,
      "learning_rate": 3.847618689476612e-06,
      "loss": 0.5296,
      "step": 2490
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.5135334730148315,
      "learning_rate": 3.7020147790418263e-06,
      "loss": 0.5256,
      "step": 2500
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.5844847559928894,
      "learning_rate": 3.5589987486958243e-06,
      "loss": 0.5379,
      "step": 2510
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.5451521277427673,
      "learning_rate": 3.418587976060653e-06,
      "loss": 0.5415,
      "step": 2520
    },
    {
      "epoch": 2.5300000000000002,
      "grad_norm": 0.5308763384819031,
      "learning_rate": 3.280799522198144e-06,
      "loss": 0.539,
      "step": 2530
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.5386905074119568,
      "learning_rate": 3.145650129536862e-06,
      "loss": 0.5291,
      "step": 2540
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.5555285215377808,
      "learning_rate": 3.013156219837776e-06,
      "loss": 0.5304,
      "step": 2550
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.5238758325576782,
      "learning_rate": 2.883333892198853e-06,
      "loss": 0.5246,
      "step": 2560
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.48604971170425415,
      "learning_rate": 2.7561989210989235e-06,
      "loss": 0.5283,
      "step": 2570
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.5265457034111023,
      "learning_rate": 2.6317667544809134e-06,
      "loss": 0.5225,
      "step": 2580
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.5299903154373169,
      "learning_rate": 2.510052511874822e-06,
      "loss": 0.5198,
      "step": 2590
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.499087929725647,
      "learning_rate": 2.391070982560564e-06,
      "loss": 0.5245,
      "step": 2600
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.4687907099723816,
      "learning_rate": 2.2748366237709374e-06,
      "loss": 0.5285,
      "step": 2610
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.5042903423309326,
      "learning_rate": 2.1613635589349756e-06,
      "loss": 0.536,
      "step": 2620
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.5207704901695251,
      "learning_rate": 2.0506655759618244e-06,
      "loss": 0.5442,
      "step": 2630
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.5210204720497131,
      "learning_rate": 1.9427561255653816e-06,
      "loss": 0.5322,
      "step": 2640
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.5390477776527405,
      "learning_rate": 1.837648319629956e-06,
      "loss": 0.5281,
      "step": 2650
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.5613718032836914,
      "learning_rate": 1.735354929617042e-06,
      "loss": 0.5423,
      "step": 2660
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.5014131665229797,
      "learning_rate": 1.6358883850134816e-06,
      "loss": 0.5373,
      "step": 2670
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.4976540803909302,
      "learning_rate": 1.5392607718211994e-06,
      "loss": 0.5281,
      "step": 2680
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.5201677083969116,
      "learning_rate": 1.4454838310886425e-06,
      "loss": 0.531,
      "step": 2690
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.4653576612472534,
      "learning_rate": 1.3545689574841342e-06,
      "loss": 0.5251,
      "step": 2700
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.5000585913658142,
      "learning_rate": 1.266527197911352e-06,
      "loss": 0.5253,
      "step": 2710
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 0.4802665114402771,
      "learning_rate": 1.1813692501670276e-06,
      "loss": 0.5328,
      "step": 2720
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.5173624157905579,
      "learning_rate": 1.0991054616410589e-06,
      "loss": 0.5339,
      "step": 2730
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.5062382221221924,
      "learning_rate": 1.0197458280592542e-06,
      "loss": 0.5368,
      "step": 2740
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.5551048517227173,
      "learning_rate": 9.432999922687396e-07,
      "loss": 0.5222,
      "step": 2750
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.5234004855155945,
      "learning_rate": 8.697772430662859e-07,
      "loss": 0.5183,
      "step": 2760
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.507086455821991,
      "learning_rate": 7.991865140696331e-07,
      "loss": 0.5204,
      "step": 2770
    },
    {
      "epoch": 2.7800000000000002,
      "grad_norm": 0.5369780659675598,
      "learning_rate": 7.315363826320005e-07,
      "loss": 0.5398,
      "step": 2780
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.5381918549537659,
      "learning_rate": 6.668350687998565e-07,
      "loss": 0.5188,
      "step": 2790
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.5035576820373535,
      "learning_rate": 6.050904343141095e-07,
      "loss": 0.5431,
      "step": 2800
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.45299527049064636,
      "learning_rate": 5.463099816548579e-07,
      "loss": 0.5319,
      "step": 2810
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.5247001647949219,
      "learning_rate": 4.905008531297661e-07,
      "loss": 0.5344,
      "step": 2820
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.4998054802417755,
      "learning_rate": 4.3766983000621266e-07,
      "loss": 0.5266,
      "step": 2830
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.5558290481567383,
      "learning_rate": 3.8782333168732033e-07,
      "loss": 0.5287,
      "step": 2840
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.5186458230018616,
      "learning_rate": 3.4096741493194197e-07,
      "loss": 0.5379,
      "step": 2850
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.499171644449234,
      "learning_rate": 2.9710777311871e-07,
      "loss": 0.5366,
      "step": 2860
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.5045514702796936,
      "learning_rate": 2.5624973555424815e-07,
      "loss": 0.5247,
      "step": 2870
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.507380485534668,
      "learning_rate": 2.1839826682562015e-07,
      "loss": 0.5349,
      "step": 2880
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.5166347622871399,
      "learning_rate": 1.8355796619708987e-07,
      "loss": 0.5346,
      "step": 2890
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.5490320920944214,
      "learning_rate": 1.517330670512629e-07,
      "loss": 0.5285,
      "step": 2900
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.5213032960891724,
      "learning_rate": 1.229274363747146e-07,
      "loss": 0.5318,
      "step": 2910
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.5020200610160828,
      "learning_rate": 9.71445742881022e-08,
      "loss": 0.5431,
      "step": 2920
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.5045921802520752,
      "learning_rate": 7.438761362087987e-08,
      "loss": 0.5287,
      "step": 2930
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.5005814433097839,
      "learning_rate": 5.4659319530636633e-08,
      "loss": 0.5288,
      "step": 2940
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.5101141333580017,
      "learning_rate": 3.796208916709565e-08,
      "loss": 0.5413,
      "step": 2950
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.4954812526702881,
      "learning_rate": 2.429795138085278e-08,
      "loss": 0.5126,
      "step": 2960
    },
    {
      "epoch": 2.9699999999999998,
      "grad_norm": 0.48134854435920715,
      "learning_rate": 1.3668566476848777e-08,
      "loss": 0.5367,
      "step": 2970
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.534488320350647,
      "learning_rate": 6.075226012636215e-09,
      "loss": 0.5266,
      "step": 2980
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.4778456687927246,
      "learning_rate": 1.5188526414244842e-09,
      "loss": 0.5479,
      "step": 2990
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5285366773605347,
      "learning_rate": 0.0,
      "loss": 0.5271,
      "step": 3000
    },
    {
      "epoch": 3.0,
      "step": 3000,
      "total_flos": 4.3556612022154035e+18,
      "train_loss": 0.6027277402877808,
      "train_runtime": 90142.7719,
      "train_samples_per_second": 2.13,
      "train_steps_per_second": 0.033
    }
  ],
  "logging_steps": 10,
  "max_steps": 3000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.3556612022154035e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}