{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.927710843373493,
  "eval_steps": 500,
  "global_step": 1030,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0963855421686747,
      "grad_norm": 0.8919573686450439,
      "learning_rate": 9.997674418116758e-05,
      "loss": 1.9774,
      "step": 10
    },
    {
      "epoch": 0.1927710843373494,
      "grad_norm": 0.910704767308804,
      "learning_rate": 9.990699835799469e-05,
      "loss": 1.6692,
      "step": 20
    },
    {
      "epoch": 0.2891566265060241,
      "grad_norm": 0.812298320227036,
      "learning_rate": 9.979082741033047e-05,
      "loss": 1.5706,
      "step": 30
    },
    {
      "epoch": 0.3855421686746988,
      "grad_norm": 0.7732472432905145,
      "learning_rate": 9.96283394041954e-05,
      "loss": 1.5094,
      "step": 40
    },
    {
      "epoch": 0.4819277108433735,
      "grad_norm": 0.8531022514130023,
      "learning_rate": 9.941968549125481e-05,
      "loss": 1.463,
      "step": 50
    },
    {
      "epoch": 0.5783132530120482,
      "grad_norm": 0.9116145183088319,
      "learning_rate": 9.916505976821263e-05,
      "loss": 1.4822,
      "step": 60
    },
    {
      "epoch": 0.6746987951807228,
      "grad_norm": 0.8879028360760914,
      "learning_rate": 9.886469909625625e-05,
      "loss": 1.4681,
      "step": 70
    },
    {
      "epoch": 0.7710843373493976,
      "grad_norm": 0.9147335792078393,
      "learning_rate": 9.851888288072053e-05,
      "loss": 1.3971,
      "step": 80
    },
    {
      "epoch": 0.8674698795180723,
      "grad_norm": 0.887118205770356,
      "learning_rate": 9.81279328111758e-05,
      "loss": 1.4472,
      "step": 90
    },
    {
      "epoch": 0.963855421686747,
      "grad_norm": 0.8531475377514153,
      "learning_rate": 9.769221256218164e-05,
      "loss": 1.4119,
      "step": 100
    },
    {
      "epoch": 1.0602409638554218,
      "grad_norm": 0.8430581950762995,
      "learning_rate": 9.721212745498494e-05,
      "loss": 1.3639,
      "step": 110
    },
    {
      "epoch": 1.1566265060240963,
      "grad_norm": 0.9779626124304455,
      "learning_rate": 9.668812408047679e-05,
      "loss": 1.3209,
      "step": 120
    },
    {
      "epoch": 1.2530120481927711,
      "grad_norm": 1.0822085428116495,
      "learning_rate": 9.612068988375897e-05,
      "loss": 1.3343,
      "step": 130
    },
    {
      "epoch": 1.3493975903614457,
      "grad_norm": 1.0414395683595772,
      "learning_rate": 9.551035271070664e-05,
      "loss": 1.2922,
      "step": 140
    },
    {
      "epoch": 1.4457831325301205,
      "grad_norm": 1.1344579456156207,
      "learning_rate": 9.485768031694872e-05,
      "loss": 1.3155,
      "step": 150
    },
    {
      "epoch": 1.5421686746987953,
      "grad_norm": 1.1349887667909784,
      "learning_rate": 9.416327983972304e-05,
      "loss": 1.2964,
      "step": 160
    },
    {
      "epoch": 1.6385542168674698,
      "grad_norm": 1.0901371692538764,
      "learning_rate": 9.342779723309745e-05,
      "loss": 1.3039,
      "step": 170
    },
    {
      "epoch": 1.7349397590361446,
      "grad_norm": 1.1237203962852838,
      "learning_rate": 9.265191666708209e-05,
      "loss": 1.3382,
      "step": 180
    },
    {
      "epoch": 1.8313253012048194,
      "grad_norm": 1.153171162807342,
      "learning_rate": 9.18363598911921e-05,
      "loss": 1.2993,
      "step": 190
    },
    {
      "epoch": 1.927710843373494,
      "grad_norm": 1.0603198764699773,
      "learning_rate": 9.098188556305263e-05,
      "loss": 1.2905,
      "step": 200
    },
    {
      "epoch": 2.0240963855421685,
      "grad_norm": 1.0860380934510327,
      "learning_rate": 9.008928854267054e-05,
      "loss": 1.2749,
      "step": 210
    },
    {
      "epoch": 2.1204819277108435,
      "grad_norm": 1.3046592302747626,
      "learning_rate": 8.915939915302968e-05,
      "loss": 1.1839,
      "step": 220
    },
    {
      "epoch": 2.216867469879518,
      "grad_norm": 1.2724868030271497,
      "learning_rate": 8.819308240769724e-05,
      "loss": 1.1675,
      "step": 230
    },
    {
      "epoch": 2.3132530120481927,
      "grad_norm": 1.4063722146992672,
      "learning_rate": 8.71912372061598e-05,
      "loss": 1.1765,
      "step": 240
    },
    {
      "epoch": 2.4096385542168672,
      "grad_norm": 1.4911158254860226,
      "learning_rate": 8.615479549763756e-05,
      "loss": 1.1846,
      "step": 250
    },
    {
      "epoch": 2.5060240963855422,
      "grad_norm": 1.4235766805578518,
      "learning_rate": 8.508472141415467e-05,
      "loss": 1.169,
      "step": 260
    },
    {
      "epoch": 2.602409638554217,
      "grad_norm": 1.38772225064988,
      "learning_rate": 8.398201037367201e-05,
      "loss": 1.1542,
      "step": 270
    },
    {
      "epoch": 2.6987951807228914,
      "grad_norm": 1.451662491739912,
      "learning_rate": 8.284768815411692e-05,
      "loss": 1.1561,
      "step": 280
    },
    {
      "epoch": 2.7951807228915664,
      "grad_norm": 1.4370300928384818,
      "learning_rate": 8.168280993917077e-05,
      "loss": 1.1738,
      "step": 290
    },
    {
      "epoch": 2.891566265060241,
      "grad_norm": 1.40886699437724,
      "learning_rate": 8.048845933670273e-05,
      "loss": 1.1636,
      "step": 300
    },
    {
      "epoch": 2.9879518072289155,
      "grad_norm": 1.4108567826491814,
      "learning_rate": 7.926574737076211e-05,
      "loss": 1.1971,
      "step": 310
    },
    {
      "epoch": 3.0843373493975905,
      "grad_norm": 1.7067948053286397,
      "learning_rate": 7.801581144806752e-05,
      "loss": 1.0335,
      "step": 320
    },
    {
      "epoch": 3.180722891566265,
      "grad_norm": 1.6598822585089195,
      "learning_rate": 7.673981429995372e-05,
      "loss": 1.0365,
      "step": 330
    },
    {
      "epoch": 3.2771084337349397,
      "grad_norm": 1.7326188759467145,
      "learning_rate": 7.543894290076103e-05,
      "loss": 1.0218,
      "step": 340
    },
    {
      "epoch": 3.3734939759036147,
      "grad_norm": 1.7737673244546652,
      "learning_rate": 7.411440736367281e-05,
      "loss": 1.0187,
      "step": 350
    },
    {
      "epoch": 3.4698795180722892,
      "grad_norm": 1.7613273642895082,
      "learning_rate": 7.276743981502856e-05,
      "loss": 1.0217,
      "step": 360
    },
    {
      "epoch": 3.566265060240964,
      "grad_norm": 1.880423633009527,
      "learning_rate": 7.139929324815965e-05,
      "loss": 1.0378,
      "step": 370
    },
    {
      "epoch": 3.662650602409639,
      "grad_norm": 1.9055198219521683,
      "learning_rate": 7.00112403578139e-05,
      "loss": 1.0406,
      "step": 380
    },
    {
      "epoch": 3.7590361445783134,
      "grad_norm": 1.926797264762729,
      "learning_rate": 6.860457235625322e-05,
      "loss": 1.0331,
      "step": 390
    },
    {
      "epoch": 3.855421686746988,
      "grad_norm": 1.9427994346863178,
      "learning_rate": 6.718059777212567e-05,
      "loss": 1.0354,
      "step": 400
    },
    {
      "epoch": 3.9518072289156625,
      "grad_norm": 1.9647182978122792,
      "learning_rate": 6.574064123322925e-05,
      "loss": 1.0445,
      "step": 410
    },
    {
      "epoch": 4.048192771084337,
      "grad_norm": 1.837995364067002,
      "learning_rate": 6.42860422342998e-05,
      "loss": 0.9521,
      "step": 420
    },
    {
      "epoch": 4.144578313253012,
      "grad_norm": 2.148688266798288,
      "learning_rate": 6.281815389096903e-05,
      "loss": 0.8772,
      "step": 430
    },
    {
      "epoch": 4.240963855421687,
      "grad_norm": 2.376205483254627,
      "learning_rate": 6.133834168105206e-05,
      "loss": 0.8616,
      "step": 440
    },
    {
      "epoch": 4.337349397590361,
      "grad_norm": 2.2364751098182527,
      "learning_rate": 5.9847982174335316e-05,
      "loss": 0.8602,
      "step": 450
    },
    {
      "epoch": 4.433734939759036,
      "grad_norm": 2.240796410509819,
      "learning_rate": 5.8348461752046116e-05,
      "loss": 0.8754,
      "step": 460
    },
    {
      "epoch": 4.530120481927711,
      "grad_norm": 2.2490246329419574,
      "learning_rate": 5.6841175317195515e-05,
      "loss": 0.8893,
      "step": 470
    },
    {
      "epoch": 4.626506024096385,
      "grad_norm": 2.328620328311832,
      "learning_rate": 5.532752499699381e-05,
      "loss": 0.8905,
      "step": 480
    },
    {
      "epoch": 4.72289156626506,
      "grad_norm": 2.4232879309790314,
      "learning_rate": 5.380891883854591e-05,
      "loss": 0.8761,
      "step": 490
    },
    {
      "epoch": 4.8192771084337345,
      "grad_norm": 2.330347626994359,
      "learning_rate": 5.228676949903973e-05,
      "loss": 0.8947,
      "step": 500
    },
    {
      "epoch": 4.9156626506024095,
      "grad_norm": 2.352453843969406,
      "learning_rate": 5.07624929316463e-05,
      "loss": 0.8807,
      "step": 510
    },
    {
      "epoch": 5.0120481927710845,
      "grad_norm": 2.2677109085585476,
      "learning_rate": 4.923750706835371e-05,
      "loss": 0.877,
      "step": 520
    },
    {
      "epoch": 5.108433734939759,
      "grad_norm": 2.929089690621969,
      "learning_rate": 4.771323050096028e-05,
      "loss": 0.6911,
      "step": 530
    },
    {
      "epoch": 5.204819277108434,
      "grad_norm": 2.6683793778473053,
      "learning_rate": 4.619108116145411e-05,
      "loss": 0.7144,
      "step": 540
    },
    {
      "epoch": 5.301204819277109,
      "grad_norm": 2.7047381812494806,
      "learning_rate": 4.46724750030062e-05,
      "loss": 0.725,
      "step": 550
    },
    {
      "epoch": 5.397590361445783,
      "grad_norm": 2.851421639172732,
      "learning_rate": 4.31588246828045e-05,
      "loss": 0.7243,
      "step": 560
    },
    {
      "epoch": 5.493975903614458,
      "grad_norm": 2.776878874955075,
      "learning_rate": 4.16515382479539e-05,
      "loss": 0.6986,
      "step": 570
    },
    {
      "epoch": 5.590361445783133,
      "grad_norm": 2.9830553324496734,
      "learning_rate": 4.015201782566471e-05,
      "loss": 0.7311,
      "step": 580
    },
    {
      "epoch": 5.686746987951807,
      "grad_norm": 2.8075065703920905,
      "learning_rate": 3.866165831894796e-05,
      "loss": 0.7511,
      "step": 590
    },
    {
      "epoch": 5.783132530120482,
      "grad_norm": 3.060856835395072,
      "learning_rate": 3.7181846109031005e-05,
      "loss": 0.7277,
      "step": 600
    },
    {
      "epoch": 5.879518072289157,
      "grad_norm": 2.9964402401459407,
      "learning_rate": 3.571395776570023e-05,
      "loss": 0.7507,
      "step": 610
    },
    {
      "epoch": 5.975903614457831,
      "grad_norm": 2.840908376303377,
      "learning_rate": 3.4259358766770766e-05,
      "loss": 0.7488,
      "step": 620
    },
    {
      "epoch": 6.072289156626506,
      "grad_norm": 3.4568397677126326,
      "learning_rate": 3.2819402227874365e-05,
      "loss": 0.6285,
      "step": 630
    },
    {
      "epoch": 6.168674698795181,
      "grad_norm": 3.1654598373448137,
      "learning_rate": 3.1395427643746796e-05,
      "loss": 0.5836,
      "step": 640
    },
    {
      "epoch": 6.265060240963855,
      "grad_norm": 3.2675864044433585,
      "learning_rate": 2.9988759642186097e-05,
      "loss": 0.5849,
      "step": 650
    },
    {
      "epoch": 6.36144578313253,
      "grad_norm": 3.2183821392920664,
      "learning_rate": 2.860070675184036e-05,
      "loss": 0.5949,
      "step": 660
    },
    {
      "epoch": 6.457831325301205,
      "grad_norm": 3.0719394770792885,
      "learning_rate": 2.7232560184971434e-05,
      "loss": 0.5786,
      "step": 670
    },
    {
      "epoch": 6.554216867469879,
      "grad_norm": 3.3384530164789523,
      "learning_rate": 2.588559263632719e-05,
      "loss": 0.5883,
      "step": 680
    },
    {
      "epoch": 6.650602409638554,
      "grad_norm": 3.257889301524805,
      "learning_rate": 2.456105709923897e-05,
      "loss": 0.5929,
      "step": 690
    },
    {
      "epoch": 6.746987951807229,
      "grad_norm": 3.3823887425006545,
      "learning_rate": 2.3260185700046294e-05,
      "loss": 0.594,
      "step": 700
    },
    {
      "epoch": 6.843373493975903,
      "grad_norm": 3.435900725711256,
      "learning_rate": 2.1984188551932512e-05,
      "loss": 0.5924,
      "step": 710
    },
    {
      "epoch": 6.9397590361445785,
      "grad_norm": 3.4675705465638473,
      "learning_rate": 2.0734252629237894e-05,
      "loss": 0.5927,
      "step": 720
    },
    {
      "epoch": 7.036144578313253,
      "grad_norm": 2.905863499929125,
      "learning_rate": 1.9511540663297285e-05,
      "loss": 0.5441,
      "step": 730
    },
    {
      "epoch": 7.132530120481928,
      "grad_norm": 3.525660740561049,
      "learning_rate": 1.831719006082924e-05,
      "loss": 0.4634,
      "step": 740
    },
    {
      "epoch": 7.228915662650603,
      "grad_norm": 3.293264937231894,
      "learning_rate": 1.7152311845883095e-05,
      "loss": 0.4671,
      "step": 750
    },
    {
      "epoch": 7.325301204819277,
      "grad_norm": 3.4643723238577904,
      "learning_rate": 1.601798962632799e-05,
      "loss": 0.476,
      "step": 760
    },
    {
      "epoch": 7.421686746987952,
      "grad_norm": 3.188250058544125,
      "learning_rate": 1.491527858584535e-05,
      "loss": 0.4981,
      "step": 770
    },
    {
      "epoch": 7.518072289156627,
      "grad_norm": 3.4201215516664516,
      "learning_rate": 1.384520450236244e-05,
      "loss": 0.4714,
      "step": 780
    },
    {
      "epoch": 7.614457831325301,
      "grad_norm": 3.2016173663444145,
      "learning_rate": 1.2808762793840201e-05,
      "loss": 0.4878,
      "step": 790
    },
    {
      "epoch": 7.710843373493976,
      "grad_norm": 3.513424701410591,
      "learning_rate": 1.1806917592302762e-05,
      "loss": 0.4914,
      "step": 800
    },
    {
      "epoch": 7.807228915662651,
      "grad_norm": 3.4642537282355543,
      "learning_rate": 1.0840600846970334e-05,
      "loss": 0.4742,
      "step": 810
    },
    {
      "epoch": 7.903614457831325,
      "grad_norm": 3.4479929396821443,
      "learning_rate": 9.91071145732948e-06,
      "loss": 0.4749,
      "step": 820
    },
    {
      "epoch": 8.0,
      "grad_norm": 3.417147882208991,
      "learning_rate": 9.018114436947373e-06,
      "loss": 0.4918,
      "step": 830
    },
    {
      "epoch": 8.096385542168674,
      "grad_norm": 3.152072043029231,
      "learning_rate": 8.163640108807896e-06,
      "loss": 0.4106,
      "step": 840
    },
    {
      "epoch": 8.19277108433735,
      "grad_norm": 3.5086623296717816,
      "learning_rate": 7.348083332917926e-06,
      "loss": 0.4141,
      "step": 850
    },
    {
      "epoch": 8.289156626506024,
      "grad_norm": 3.727099779414922,
      "learning_rate": 6.572202766902569e-06,
      "loss": 0.4287,
      "step": 860
    },
    {
      "epoch": 8.385542168674698,
      "grad_norm": 3.356502113157289,
      "learning_rate": 5.83672016027697e-06,
      "loss": 0.3974,
      "step": 870
    },
    {
      "epoch": 8.481927710843374,
      "grad_norm": 3.5114290988775427,
      "learning_rate": 5.1423196830513e-06,
      "loss": 0.4111,
      "step": 880
    },
    {
      "epoch": 8.578313253012048,
      "grad_norm": 3.3470018501511447,
      "learning_rate": 4.489647289293369e-06,
      "loss": 0.4165,
      "step": 890
    },
    {
      "epoch": 8.674698795180722,
      "grad_norm": 3.1455647605463666,
      "learning_rate": 3.879310116241042e-06,
      "loss": 0.4156,
      "step": 900
    },
    {
      "epoch": 8.771084337349398,
      "grad_norm": 3.4300111085823124,
      "learning_rate": 3.3118759195232275e-06,
      "loss": 0.3935,
      "step": 910
    },
    {
      "epoch": 8.867469879518072,
      "grad_norm": 3.2501800712068825,
      "learning_rate": 2.787872545015069e-06,
      "loss": 0.4103,
      "step": 920
    },
    {
      "epoch": 8.963855421686747,
      "grad_norm": 3.2829723029519666,
      "learning_rate": 2.307787437818365e-06,
      "loss": 0.4131,
      "step": 930
    },
    {
      "epoch": 9.060240963855422,
      "grad_norm": 3.145951812862308,
      "learning_rate": 1.8720671888242059e-06,
      "loss": 0.3828,
      "step": 940
    },
    {
      "epoch": 9.156626506024097,
      "grad_norm": 3.1956957268186112,
      "learning_rate": 1.4811171192794627e-06,
      "loss": 0.365,
      "step": 950
    },
    {
      "epoch": 9.25301204819277,
      "grad_norm": 3.024863424355031,
      "learning_rate": 1.1353009037437523e-06,
      "loss": 0.3736,
      "step": 960
    },
    {
      "epoch": 9.349397590361447,
      "grad_norm": 3.4426367987042106,
      "learning_rate": 8.349402317873789e-07,
      "loss": 0.378,
      "step": 970
    },
    {
      "epoch": 9.44578313253012,
      "grad_norm": 3.3566957382084515,
      "learning_rate": 5.803145087451945e-07,
      "loss": 0.4013,
      "step": 980
    },
    {
      "epoch": 9.542168674698795,
      "grad_norm": 3.616442803141579,
      "learning_rate": 3.716605958046071e-07,
      "loss": 0.386,
      "step": 990
    },
    {
      "epoch": 9.638554216867469,
      "grad_norm": 3.1110352724634835,
      "learning_rate": 2.0917258966953733e-07,
      "loss": 0.3862,
      "step": 1000
    },
    {
      "epoch": 9.734939759036145,
      "grad_norm": 3.3766596212586504,
      "learning_rate": 9.300164200530814e-08,
      "loss": 0.3808,
      "step": 1010
    },
    {
      "epoch": 9.831325301204819,
      "grad_norm": 3.202391216065192,
      "learning_rate": 2.3255818832423894e-08,
      "loss": 0.3844,
      "step": 1020
    },
    {
      "epoch": 9.927710843373493,
      "grad_norm": 3.3375553267537876,
      "learning_rate": 0.0,
      "loss": 0.367,
      "step": 1030
    },
    {
      "epoch": 9.927710843373493,
      "step": 1030,
      "total_flos": 144813601193984.0,
      "train_loss": 0.8549524587334938,
      "train_runtime": 9342.774,
      "train_samples_per_second": 0.888,
      "train_steps_per_second": 0.11
    }
  ],
  "logging_steps": 10,
  "max_steps": 1030,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 144813601193984.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}