{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2,
  "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": 0.16093474626541138,
      "learning_rate": 3.6e-06,
      "loss": 1.2616,
      "step": 10
    },
    {
      "grad_norm": 0.146929532289505,
      "learning_rate": 7.6e-06,
      "loss": 1.2639,
      "step": 20
    },
    {
      "grad_norm": 0.20482921600341797,
      "learning_rate": 1.16e-05,
      "loss": 1.253,
      "step": 30
    },
    {
      "grad_norm": 0.289375364780426,
      "learning_rate": 1.56e-05,
      "loss": 1.2107,
      "step": 40
    },
    {
      "grad_norm": 0.1968756765127182,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 1.1575,
      "step": 50
    },
    {
      "grad_norm": 0.1628505289554596,
      "learning_rate": 2.36e-05,
      "loss": 1.1278,
      "step": 60
    },
    {
      "grad_norm": 0.3194240629673004,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 1.1057,
      "step": 70
    },
    {
      "grad_norm": 0.1577083170413971,
      "learning_rate": 3.16e-05,
      "loss": 1.0875,
      "step": 80
    },
    {
      "grad_norm": 0.3057079017162323,
      "learning_rate": 3.56e-05,
      "loss": 1.0584,
      "step": 90
    },
    {
      "grad_norm": 0.5276331901550293,
      "learning_rate": 3.960000000000001e-05,
      "loss": 1.0145,
      "step": 100
    },
    {
      "grad_norm": 0.37284791469573975,
      "learning_rate": 4.36e-05,
      "loss": 0.9744,
      "step": 110
    },
    {
      "grad_norm": 0.3288450837135315,
      "learning_rate": 4.76e-05,
      "loss": 0.9222,
      "step": 120
    },
    {
      "grad_norm": 1.0112025737762451,
      "learning_rate": 5.16e-05,
      "loss": 0.8713,
      "step": 130
    },
    {
      "grad_norm": 0.5174235105514526,
      "learning_rate": 5.560000000000001e-05,
      "loss": 0.8248,
      "step": 140
    },
    {
      "grad_norm": 0.45569029450416565,
      "learning_rate": 5.96e-05,
      "loss": 0.7778,
      "step": 150
    },
    {
      "grad_norm": 0.7311891913414001,
      "learning_rate": 6.36e-05,
      "loss": 0.733,
      "step": 160
    },
    {
      "grad_norm": 0.5297268629074097,
      "learning_rate": 6.76e-05,
      "loss": 0.6964,
      "step": 170
    },
    {
      "grad_norm": 0.40571925044059753,
      "learning_rate": 7.16e-05,
      "loss": 0.6587,
      "step": 180
    },
    {
      "grad_norm": 0.8430274128913879,
      "learning_rate": 7.560000000000001e-05,
      "loss": 0.6186,
      "step": 190
    },
    {
      "grad_norm": 0.5413951277732849,
      "learning_rate": 7.960000000000001e-05,
      "loss": 0.5931,
      "step": 200
    },
    {
      "grad_norm": 0.5479915738105774,
      "learning_rate": 8.36e-05,
      "loss": 0.5662,
      "step": 210
    },
    {
      "grad_norm": 0.5007767677307129,
      "learning_rate": 8.76e-05,
      "loss": 0.5469,
      "step": 220
    },
    {
      "grad_norm": 0.5148480534553528,
      "learning_rate": 9.16e-05,
      "loss": 0.5208,
      "step": 230
    },
    {
      "grad_norm": 0.6750017404556274,
      "learning_rate": 9.56e-05,
      "loss": 0.5032,
      "step": 240
    },
    {
      "grad_norm": 0.6098655462265015,
      "learning_rate": 9.960000000000001e-05,
      "loss": 0.489,
      "step": 250
    },
    {
      "grad_norm": 0.7122275233268738,
      "learning_rate": 9.999911419878559e-05,
      "loss": 0.4701,
      "step": 260
    },
    {
      "grad_norm": 0.7136951088905334,
      "learning_rate": 9.999605221019081e-05,
      "loss": 0.4594,
      "step": 270
    },
    {
      "grad_norm": 0.7450923323631287,
      "learning_rate": 9.999080323230761e-05,
      "loss": 0.4393,
      "step": 280
    },
    {
      "grad_norm": 0.5781692266464233,
      "learning_rate": 9.998336749474329e-05,
      "loss": 0.4276,
      "step": 290
    },
    {
      "grad_norm": 0.7031801342964172,
      "learning_rate": 9.997374532276107e-05,
      "loss": 0.4148,
      "step": 300
    },
    {
      "grad_norm": 0.8787388205528259,
      "learning_rate": 9.996193713726596e-05,
      "loss": 0.3996,
      "step": 310
    },
    {
      "grad_norm": 0.682303249835968,
      "learning_rate": 9.994794345478624e-05,
      "loss": 0.3901,
      "step": 320
    },
    {
      "grad_norm": 0.4642148017883301,
      "learning_rate": 9.99317648874509e-05,
      "loss": 0.3758,
      "step": 330
    },
    {
      "grad_norm": 0.48445796966552734,
      "learning_rate": 9.991340214296292e-05,
      "loss": 0.3576,
      "step": 340
    },
    {
      "grad_norm": 0.6353476047515869,
      "learning_rate": 9.989285602456819e-05,
      "loss": 0.3465,
      "step": 350
    },
    {
      "grad_norm": 0.7517806887626648,
      "learning_rate": 9.98701274310205e-05,
      "loss": 0.3401,
      "step": 360
    },
    {
      "grad_norm": 0.7166535258293152,
      "learning_rate": 9.984521735654218e-05,
      "loss": 0.3358,
      "step": 370
    },
    {
      "grad_norm": 0.5661184787750244,
      "learning_rate": 9.981812689078057e-05,
      "loss": 0.3265,
      "step": 380
    },
    {
      "grad_norm": 0.576542854309082,
      "learning_rate": 9.978885721876041e-05,
      "loss": 0.319,
      "step": 390
    },
    {
      "grad_norm": 0.6311920881271362,
      "learning_rate": 9.975740962083198e-05,
      "loss": 0.3079,
      "step": 400
    },
    {
      "grad_norm": 0.6936492919921875,
      "learning_rate": 9.972378547261504e-05,
      "loss": 0.2968,
      "step": 410
    },
    {
      "grad_norm": 0.6447014212608337,
      "learning_rate": 9.968798624493885e-05,
      "loss": 0.2936,
      "step": 420
    },
    {
      "grad_norm": 0.6125183701515198,
      "learning_rate": 9.965001350377753e-05,
      "loss": 0.2874,
      "step": 430
    },
    {
      "grad_norm": 0.5859553813934326,
      "learning_rate": 9.960986891018183e-05,
      "loss": 0.2784,
      "step": 440
    },
    {
      "grad_norm": 0.76313716173172,
      "learning_rate": 9.95675542202063e-05,
      "loss": 0.2743,
      "step": 450
    },
    {
      "grad_norm": 0.7491763234138489,
      "learning_rate": 9.952307128483256e-05,
      "loss": 0.2594,
      "step": 460
    },
    {
      "grad_norm": 0.565811276435852,
      "learning_rate": 9.947642204988835e-05,
      "loss": 0.252,
      "step": 470
    },
    {
      "grad_norm": 0.881474494934082,
      "learning_rate": 9.942760855596226e-05,
      "loss": 0.2432,
      "step": 480
    },
    {
      "grad_norm": 0.5126052498817444,
      "learning_rate": 9.937663293831471e-05,
      "loss": 0.2302,
      "step": 490
    },
    {
      "grad_norm": 0.5917734503746033,
      "learning_rate": 9.932349742678433e-05,
      "loss": 0.2232,
      "step": 500
    },
    {
      "grad_norm": 0.5307651162147522,
      "learning_rate": 9.926820434569051e-05,
      "loss": 0.2133,
      "step": 510
    },
    {
      "grad_norm": 0.7390123605728149,
      "learning_rate": 9.921075611373179e-05,
      "loss": 0.1937,
      "step": 520
    },
    {
      "grad_norm": 0.7391104102134705,
      "learning_rate": 9.915115524387988e-05,
      "loss": 0.1818,
      "step": 530
    },
    {
      "grad_norm": 0.7110467553138733,
      "learning_rate": 9.908940434326997e-05,
      "loss": 0.1845,
      "step": 540
    },
    {
      "grad_norm": 0.679317057132721,
      "learning_rate": 9.902550611308645e-05,
      "loss": 0.1654,
      "step": 550
    },
    {
      "grad_norm": 0.4592258036136627,
      "learning_rate": 9.895946334844494e-05,
      "loss": 0.1591,
      "step": 560
    },
    {
      "grad_norm": 0.7367031574249268,
      "learning_rate": 9.889127893826989e-05,
      "loss": 0.1586,
      "step": 570
    },
    {
      "grad_norm": 0.7603464722633362,
      "learning_rate": 9.882095586516831e-05,
      "loss": 0.1465,
      "step": 580
    },
    {
      "grad_norm": 0.6639998555183411,
      "learning_rate": 9.874849720529921e-05,
      "loss": 0.1458,
      "step": 590
    },
    {
      "grad_norm": 0.6239601969718933,
      "learning_rate": 9.867390612823914e-05,
      "loss": 0.1402,
      "step": 600
    },
    {
      "grad_norm": 0.634850800037384,
      "learning_rate": 9.859718589684344e-05,
      "loss": 0.1343,
      "step": 610
    },
    {
      "grad_norm": 0.6103629469871521,
      "learning_rate": 9.851833986710353e-05,
      "loss": 0.1346,
      "step": 620
    },
    {
      "grad_norm": 0.6501741409301758,
      "learning_rate": 9.843737148800023e-05,
      "loss": 0.1271,
      "step": 630
    },
    {
      "grad_norm": 0.5558211207389832,
      "learning_rate": 9.835428430135271e-05,
      "loss": 0.1298,
      "step": 640
    },
    {
      "grad_norm": 0.5191821455955505,
      "learning_rate": 9.82690819416637e-05,
      "loss": 0.1264,
      "step": 650
    },
    {
      "grad_norm": 0.5054128170013428,
      "learning_rate": 9.818176813596041e-05,
      "loss": 0.1146,
      "step": 660
    },
    {
      "grad_norm": 0.5061753392219543,
      "learning_rate": 9.809234670363159e-05,
      "loss": 0.1287,
      "step": 670
    },
    {
      "grad_norm": 0.5858749151229858,
      "learning_rate": 9.800082155626034e-05,
      "loss": 0.124,
      "step": 680
    },
    {
      "grad_norm": 0.5644469261169434,
      "learning_rate": 9.790719669745312e-05,
      "loss": 0.12,
      "step": 690
    },
    {
      "grad_norm": 0.6109688878059387,
      "learning_rate": 9.781147622266455e-05,
      "loss": 0.12,
      "step": 700
    },
    {
      "grad_norm": 0.5990709662437439,
      "learning_rate": 9.771366431901831e-05,
      "loss": 0.1155,
      "step": 710
    },
    {
      "grad_norm": 0.4833641052246094,
      "learning_rate": 9.761376526512394e-05,
      "loss": 0.1118,
      "step": 720
    },
    {
      "grad_norm": 0.5707718133926392,
      "learning_rate": 9.751178343088963e-05,
      "loss": 0.1224,
      "step": 730
    },
    {
      "grad_norm": 0.5550230145454407,
      "learning_rate": 9.740772327733123e-05,
      "loss": 0.1155,
      "step": 740
    },
    {
      "grad_norm": 0.6085163950920105,
      "learning_rate": 9.730158935637697e-05,
      "loss": 0.1147,
      "step": 750
    },
    {
      "grad_norm": 0.45861753821372986,
      "learning_rate": 9.719338631066834e-05,
      "loss": 0.1115,
      "step": 760
    },
    {
      "grad_norm": 0.587456464767456,
      "learning_rate": 9.708311887335713e-05,
      "loss": 0.1184,
      "step": 770
    },
    {
      "grad_norm": 0.5577616095542908,
      "learning_rate": 9.697079186789823e-05,
      "loss": 0.1172,
      "step": 780
    },
    {
      "grad_norm": 0.4791278541088104,
      "learning_rate": 9.685641020783876e-05,
      "loss": 0.1145,
      "step": 790
    },
    {
      "grad_norm": 0.5386468768119812,
      "learning_rate": 9.67399788966031e-05,
      "loss": 0.1149,
      "step": 800
    },
    {
      "grad_norm": 0.5449627637863159,
      "learning_rate": 9.662150302727395e-05,
      "loss": 0.1063,
      "step": 810
    },
    {
      "grad_norm": 0.5971572399139404,
      "learning_rate": 9.650098778236968e-05,
      "loss": 0.108,
      "step": 820
    },
    {
      "grad_norm": 0.5777544975280762,
      "learning_rate": 9.637843843361749e-05,
      "loss": 0.1053,
      "step": 830
    },
    {
      "grad_norm": 0.5564993619918823,
      "learning_rate": 9.62538603417229e-05,
      "loss": 0.1094,
      "step": 840
    },
    {
      "grad_norm": 0.560504138469696,
      "learning_rate": 9.612725895613526e-05,
      "loss": 0.1175,
      "step": 850
    },
    {
      "grad_norm": 0.447893887758255,
      "learning_rate": 9.599863981480926e-05,
      "loss": 0.1082,
      "step": 860
    },
    {
      "grad_norm": 0.5009526610374451,
      "learning_rate": 9.586800854396283e-05,
      "loss": 0.1041,
      "step": 870
    },
    {
      "grad_norm": 0.3550075590610504,
      "learning_rate": 9.573537085783095e-05,
      "loss": 0.1085,
      "step": 880
    },
    {
      "grad_norm": 0.5879306793212891,
      "learning_rate": 9.560073255841571e-05,
      "loss": 0.1089,
      "step": 890
    },
    {
      "grad_norm": 0.42884793877601624,
      "learning_rate": 9.546409953523247e-05,
      "loss": 0.1099,
      "step": 900
    },
    {
      "grad_norm": 0.4561026394367218,
      "learning_rate": 9.532547776505229e-05,
      "loss": 0.1036,
      "step": 910
    },
    {
      "grad_norm": 0.4620613157749176,
      "learning_rate": 9.518487331164048e-05,
      "loss": 0.1048,
      "step": 920
    },
    {
      "grad_norm": 0.5412100553512573,
      "learning_rate": 9.504229232549134e-05,
      "loss": 0.1059,
      "step": 930
    },
    {
      "grad_norm": 0.4077288508415222,
      "learning_rate": 9.489774104355909e-05,
      "loss": 0.1005,
      "step": 940
    },
    {
      "grad_norm": 0.5104166865348816,
      "learning_rate": 9.475122578898507e-05,
      "loss": 0.1039,
      "step": 950
    },
    {
      "grad_norm": 0.457266628742218,
      "learning_rate": 9.460275297082119e-05,
      "loss": 0.1019,
      "step": 960
    },
    {
      "grad_norm": 0.46367529034614563,
      "learning_rate": 9.445232908374948e-05,
      "loss": 0.1035,
      "step": 970
    },
    {
      "grad_norm": 0.5154226422309875,
      "learning_rate": 9.429996070779808e-05,
      "loss": 0.1022,
      "step": 980
    },
    {
      "grad_norm": 0.4712366759777069,
      "learning_rate": 9.414565450805333e-05,
      "loss": 0.1067,
      "step": 990
    },
    {
      "grad_norm": 0.4741515517234802,
      "learning_rate": 9.398941723436831e-05,
      "loss": 0.1031,
      "step": 1000
    }
  ],
  "logging_steps": 10,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 1000,
  "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": 128,
  "trial_name": null,
  "trial_params": null
}