{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.3333333333333333,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "grad_norm": 0.16593393683433533,
      "learning_rate": 6e-06,
      "loss": 1.2019,
      "step": 10
    },
    {
      "grad_norm": 0.2364676296710968,
      "learning_rate": 1.2666666666666668e-05,
      "loss": 1.2001,
      "step": 20
    },
    {
      "grad_norm": 0.3223605155944824,
      "learning_rate": 1.9333333333333333e-05,
      "loss": 1.1652,
      "step": 30
    },
    {
      "grad_norm": 0.288097620010376,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 1.1203,
      "step": 40
    },
    {
      "grad_norm": 0.2957666516304016,
      "learning_rate": 3.266666666666667e-05,
      "loss": 1.1073,
      "step": 50
    },
    {
      "grad_norm": 0.3105441927909851,
      "learning_rate": 3.933333333333333e-05,
      "loss": 1.0981,
      "step": 60
    },
    {
      "grad_norm": 0.2903534770011902,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.0992,
      "step": 70
    },
    {
      "grad_norm": 0.2657025158405304,
      "learning_rate": 5.266666666666666e-05,
      "loss": 1.0949,
      "step": 80
    },
    {
      "grad_norm": 0.22208809852600098,
      "learning_rate": 5.9333333333333343e-05,
      "loss": 1.0883,
      "step": 90
    },
    {
      "grad_norm": 0.32055118680000305,
      "learning_rate": 6.6e-05,
      "loss": 1.0812,
      "step": 100
    },
    {
      "grad_norm": 0.2965445816516876,
      "learning_rate": 7.266666666666667e-05,
      "loss": 1.0723,
      "step": 110
    },
    {
      "grad_norm": 0.3756393790245056,
      "learning_rate": 7.933333333333334e-05,
      "loss": 1.0464,
      "step": 120
    },
    {
      "grad_norm": 0.5002507567405701,
      "learning_rate": 8.6e-05,
      "loss": 1.0023,
      "step": 130
    },
    {
      "grad_norm": 0.4813876152038574,
      "learning_rate": 9.266666666666666e-05,
      "loss": 0.9625,
      "step": 140
    },
    {
      "grad_norm": 0.513647198677063,
      "learning_rate": 9.933333333333334e-05,
      "loss": 0.9308,
      "step": 150
    },
    {
      "grad_norm": 0.8362369537353516,
      "learning_rate": 9.999753945398704e-05,
      "loss": 0.8938,
      "step": 160
    },
    {
      "grad_norm": 0.5613930821418762,
      "learning_rate": 9.998903417374228e-05,
      "loss": 0.862,
      "step": 170
    },
    {
      "grad_norm": 0.5763223767280579,
      "learning_rate": 9.997445481536973e-05,
      "loss": 0.8221,
      "step": 180
    },
    {
      "grad_norm": 0.5941640734672546,
      "learning_rate": 9.995380315038119e-05,
      "loss": 0.7922,
      "step": 190
    },
    {
      "grad_norm": 0.7477675080299377,
      "learning_rate": 9.99270816881235e-05,
      "loss": 0.7702,
      "step": 200
    },
    {
      "grad_norm": 0.5304926633834839,
      "learning_rate": 9.989429367547377e-05,
      "loss": 0.7553,
      "step": 210
    },
    {
      "grad_norm": 0.5700099468231201,
      "learning_rate": 9.985544309644475e-05,
      "loss": 0.7369,
      "step": 220
    },
    {
      "grad_norm": 0.7928358316421509,
      "learning_rate": 9.98105346717008e-05,
      "loss": 0.7222,
      "step": 230
    },
    {
      "grad_norm": 0.7667017579078674,
      "learning_rate": 9.97595738579843e-05,
      "loss": 0.7114,
      "step": 240
    },
    {
      "grad_norm": 0.5060045123100281,
      "learning_rate": 9.970256684745258e-05,
      "loss": 0.6861,
      "step": 250
    },
    {
      "grad_norm": 0.6732467412948608,
      "learning_rate": 9.963952056692549e-05,
      "loss": 0.675,
      "step": 260
    },
    {
      "grad_norm": 0.6021248698234558,
      "learning_rate": 9.957044267704384e-05,
      "loss": 0.6767,
      "step": 270
    },
    {
      "grad_norm": 0.6704761981964111,
      "learning_rate": 9.949534157133844e-05,
      "loss": 0.6512,
      "step": 280
    },
    {
      "grad_norm": 0.7285646200180054,
      "learning_rate": 9.941422637521035e-05,
      "loss": 0.6512,
      "step": 290
    },
    {
      "grad_norm": 0.7110922336578369,
      "learning_rate": 9.932710694482191e-05,
      "loss": 0.6196,
      "step": 300
    },
    {
      "grad_norm": 0.6850391626358032,
      "learning_rate": 9.923399386589933e-05,
      "loss": 0.6118,
      "step": 310
    },
    {
      "grad_norm": 0.6997315883636475,
      "learning_rate": 9.913489845244626e-05,
      "loss": 0.5941,
      "step": 320
    },
    {
      "grad_norm": 0.760143518447876,
      "learning_rate": 9.902983274536912e-05,
      "loss": 0.5891,
      "step": 330
    },
    {
      "grad_norm": 0.759657084941864,
      "learning_rate": 9.891880951101407e-05,
      "loss": 0.5812,
      "step": 340
    },
    {
      "grad_norm": 0.8276827931404114,
      "learning_rate": 9.880184223961573e-05,
      "loss": 0.5674,
      "step": 350
    },
    {
      "grad_norm": 0.5555763244628906,
      "learning_rate": 9.867894514365802e-05,
      "loss": 0.5544,
      "step": 360
    },
    {
      "grad_norm": 0.6141887307167053,
      "learning_rate": 9.855013315614725e-05,
      "loss": 0.5504,
      "step": 370
    },
    {
      "grad_norm": 0.7782378792762756,
      "learning_rate": 9.841542192879762e-05,
      "loss": 0.5447,
      "step": 380
    },
    {
      "grad_norm": 0.8065223693847656,
      "learning_rate": 9.82748278301294e-05,
      "loss": 0.5435,
      "step": 390
    },
    {
      "grad_norm": 0.7179561257362366,
      "learning_rate": 9.812836794348004e-05,
      "loss": 0.5344,
      "step": 400
    },
    {
      "grad_norm": 0.7837668657302856,
      "learning_rate": 9.797606006492841e-05,
      "loss": 0.5249,
      "step": 410
    },
    {
      "grad_norm": 0.7675067782402039,
      "learning_rate": 9.781792270113241e-05,
      "loss": 0.5199,
      "step": 420
    },
    {
      "grad_norm": 0.8279472589492798,
      "learning_rate": 9.765397506708023e-05,
      "loss": 0.5071,
      "step": 430
    },
    {
      "grad_norm": 0.8260670304298401,
      "learning_rate": 9.748423708375563e-05,
      "loss": 0.4926,
      "step": 440
    },
    {
      "grad_norm": 0.881014347076416,
      "learning_rate": 9.730872937571739e-05,
      "loss": 0.4907,
      "step": 450
    },
    {
      "grad_norm": 0.9371006488800049,
      "learning_rate": 9.712747326859315e-05,
      "loss": 0.4815,
      "step": 460
    },
    {
      "grad_norm": 0.8164715766906738,
      "learning_rate": 9.69404907864883e-05,
      "loss": 0.4755,
      "step": 470
    },
    {
      "grad_norm": 0.7007074952125549,
      "learning_rate": 9.674780464930979e-05,
      "loss": 0.4649,
      "step": 480
    },
    {
      "grad_norm": 0.9710970520973206,
      "learning_rate": 9.654943827000548e-05,
      "loss": 0.4578,
      "step": 490
    },
    {
      "grad_norm": 0.7813271284103394,
      "learning_rate": 9.634541575171929e-05,
      "loss": 0.4481,
      "step": 500
    },
    {
      "grad_norm": 0.8395043611526489,
      "learning_rate": 9.613576188486253e-05,
      "loss": 0.4437,
      "step": 510
    },
    {
      "grad_norm": 0.8199443221092224,
      "learning_rate": 9.59205021441015e-05,
      "loss": 0.4302,
      "step": 520
    },
    {
      "grad_norm": 0.7755788564682007,
      "learning_rate": 9.569966268526232e-05,
      "loss": 0.4159,
      "step": 530
    },
    {
      "grad_norm": 0.8591013550758362,
      "learning_rate": 9.54732703421526e-05,
      "loss": 0.3954,
      "step": 540
    },
    {
      "grad_norm": 0.8200550079345703,
      "learning_rate": 9.524135262330098e-05,
      "loss": 0.3959,
      "step": 550
    },
    {
      "grad_norm": 0.8531577587127686,
      "learning_rate": 9.50039377086147e-05,
      "loss": 0.3763,
      "step": 560
    },
    {
      "grad_norm": 0.7548572421073914,
      "learning_rate": 9.476105444595534e-05,
      "loss": 0.3647,
      "step": 570
    },
    {
      "grad_norm": 0.8562734723091125,
      "learning_rate": 9.451273234763371e-05,
      "loss": 0.3526,
      "step": 580
    },
    {
      "grad_norm": 1.0484408140182495,
      "learning_rate": 9.425900158682385e-05,
      "loss": 0.3541,
      "step": 590
    },
    {
      "grad_norm": 1.114036202430725,
      "learning_rate": 9.399989299389661e-05,
      "loss": 0.3495,
      "step": 600
    },
    {
      "grad_norm": 1.0067291259765625,
      "learning_rate": 9.373543805267368e-05,
      "loss": 0.3446,
      "step": 610
    },
    {
      "grad_norm": 1.0836315155029297,
      "learning_rate": 9.346566889660193e-05,
      "loss": 0.3169,
      "step": 620
    },
    {
      "grad_norm": 1.1337170600891113,
      "learning_rate": 9.319061830484898e-05,
      "loss": 0.3029,
      "step": 630
    },
    {
      "grad_norm": 1.0124084949493408,
      "learning_rate": 9.291031969832026e-05,
      "loss": 0.3035,
      "step": 640
    },
    {
      "grad_norm": 1.1703892946243286,
      "learning_rate": 9.262480713559808e-05,
      "loss": 0.3065,
      "step": 650
    },
    {
      "grad_norm": 0.8515862226486206,
      "learning_rate": 9.233411530880326e-05,
      "loss": 0.2916,
      "step": 660
    },
    {
      "grad_norm": 1.3676186800003052,
      "learning_rate": 9.20382795393797e-05,
      "loss": 0.2933,
      "step": 670
    },
    {
      "grad_norm": 0.8166829943656921,
      "learning_rate": 9.173733577380258e-05,
      "loss": 0.2724,
      "step": 680
    },
    {
      "grad_norm": 1.028581142425537,
      "learning_rate": 9.143132057921058e-05,
      "loss": 0.26,
      "step": 690
    },
    {
      "grad_norm": 1.035443663597107,
      "learning_rate": 9.112027113896262e-05,
      "loss": 0.2551,
      "step": 700
    },
    {
      "grad_norm": 1.22231924533844,
      "learning_rate": 9.080422524811982e-05,
      "loss": 0.2613,
      "step": 710
    },
    {
      "grad_norm": 1.0958259105682373,
      "learning_rate": 9.048322130885305e-05,
      "loss": 0.2384,
      "step": 720
    },
    {
      "grad_norm": 0.8523461222648621,
      "learning_rate": 9.015729832577681e-05,
      "loss": 0.2329,
      "step": 730
    },
    {
      "grad_norm": 0.964102029800415,
      "learning_rate": 8.982649590120982e-05,
      "loss": 0.2094,
      "step": 740
    },
    {
      "grad_norm": 0.98223876953125,
      "learning_rate": 8.949085423036296e-05,
      "loss": 0.2132,
      "step": 750
    },
    {
      "grad_norm": 1.0459887981414795,
      "learning_rate": 8.91504140964553e-05,
      "loss": 0.1981,
      "step": 760
    },
    {
      "grad_norm": 1.3035297393798828,
      "learning_rate": 8.880521686575857e-05,
      "loss": 0.1954,
      "step": 770
    },
    {
      "grad_norm": 0.9774158596992493,
      "learning_rate": 8.845530448257085e-05,
      "loss": 0.1891,
      "step": 780
    },
    {
      "grad_norm": 1.1295584440231323,
      "learning_rate": 8.810071946411989e-05,
      "loss": 0.1783,
      "step": 790
    },
    {
      "grad_norm": 0.8923147916793823,
      "learning_rate": 8.774150489539707e-05,
      "loss": 0.1765,
      "step": 800
    },
    {
      "grad_norm": 1.3010443449020386,
      "learning_rate": 8.737770442392212e-05,
      "loss": 0.1742,
      "step": 810
    },
    {
      "grad_norm": 0.9801855683326721,
      "learning_rate": 8.700936225443959e-05,
      "loss": 0.1587,
      "step": 820
    },
    {
      "grad_norm": 1.0634956359863281,
      "learning_rate": 8.663652314354765e-05,
      "loss": 0.1646,
      "step": 830
    },
    {
      "grad_norm": 0.8811051845550537,
      "learning_rate": 8.625923239425978e-05,
      "loss": 0.1485,
      "step": 840
    },
    {
      "grad_norm": 1.0639090538024902,
      "learning_rate": 8.587753585050004e-05,
      "loss": 0.1585,
      "step": 850
    },
    {
      "grad_norm": 0.9883255362510681,
      "learning_rate": 8.549147989153276e-05,
      "loss": 0.1443,
      "step": 860
    },
    {
      "grad_norm": 1.4074150323867798,
      "learning_rate": 8.510111142632698e-05,
      "loss": 0.1388,
      "step": 870
    },
    {
      "grad_norm": 1.2396678924560547,
      "learning_rate": 8.470647788785665e-05,
      "loss": 0.1283,
      "step": 880
    },
    {
      "grad_norm": 1.1257601976394653,
      "learning_rate": 8.430762722733714e-05,
      "loss": 0.1183,
      "step": 890
    },
    {
      "grad_norm": 0.95068359375,
      "learning_rate": 8.390460790839882e-05,
      "loss": 0.1372,
      "step": 900
    },
    {
      "grad_norm": 1.1750320196151733,
      "learning_rate": 8.349746890119826e-05,
      "loss": 0.1125,
      "step": 910
    },
    {
      "grad_norm": 1.0115416049957275,
      "learning_rate": 8.308625967646795e-05,
      "loss": 0.1298,
      "step": 920
    },
    {
      "grad_norm": 0.9820786118507385,
      "learning_rate": 8.267103019950529e-05,
      "loss": 0.1199,
      "step": 930
    },
    {
      "grad_norm": 0.9424199461936951,
      "learning_rate": 8.225183092410128e-05,
      "loss": 0.1097,
      "step": 940
    },
    {
      "grad_norm": 0.804672122001648,
      "learning_rate": 8.182871278641009e-05,
      "loss": 0.1264,
      "step": 950
    },
    {
      "grad_norm": 1.0112346410751343,
      "learning_rate": 8.140172719875979e-05,
      "loss": 0.107,
      "step": 960
    },
    {
      "grad_norm": 0.9198870062828064,
      "learning_rate": 8.097092604340542e-05,
      "loss": 0.1114,
      "step": 970
    },
    {
      "grad_norm": 0.8466675877571106,
      "learning_rate": 8.053636166622476e-05,
      "loss": 0.1178,
      "step": 980
    },
    {
      "grad_norm": 0.9316571354866028,
      "learning_rate": 8.009808687035798e-05,
      "loss": 0.1163,
      "step": 990
    },
    {
      "grad_norm": 1.0110657215118408,
      "learning_rate": 7.965615490979163e-05,
      "loss": 0.1149,
      "step": 1000
    }
  ],
  "logging_steps": 10,
  "max_steps": 3000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}