{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.18898629849335924,
  "eval_steps": 500,
  "global_step": 900,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002099847761037325,
      "grad_norm": 22925.884765625,
      "learning_rate": 1.9999999999999996e-07,
      "loss": 9.5004,
      "step": 10
    },
    {
      "epoch": 0.00419969552207465,
      "grad_norm": 1589.221435546875,
      "learning_rate": 1.2e-06,
      "loss": 8.7173,
      "step": 20
    },
    {
      "epoch": 0.006299543283111974,
      "grad_norm": 577.2943115234375,
      "learning_rate": 2.1999999999999997e-06,
      "loss": 7.6284,
      "step": 30
    },
    {
      "epoch": 0.0083993910441493,
      "grad_norm": 308.8800048828125,
      "learning_rate": 3.1999999999999994e-06,
      "loss": 7.2988,
      "step": 40
    },
    {
      "epoch": 0.010499238805186623,
      "grad_norm": 458.5738220214844,
      "learning_rate": 4.2e-06,
      "loss": 7.1397,
      "step": 50
    },
    {
      "epoch": 0.012599086566223949,
      "grad_norm": 2117.726806640625,
      "learning_rate": 5.199999999999999e-06,
      "loss": 7.0424,
      "step": 60
    },
    {
      "epoch": 0.014698934327261274,
      "grad_norm": 784.7866821289062,
      "learning_rate": 6.199999999999999e-06,
      "loss": 6.9896,
      "step": 70
    },
    {
      "epoch": 0.0167987820882986,
      "grad_norm": 719.2965087890625,
      "learning_rate": 7.2e-06,
      "loss": 6.9514,
      "step": 80
    },
    {
      "epoch": 0.018898629849335925,
      "grad_norm": 887.7239379882812,
      "learning_rate": 8.2e-06,
      "loss": 6.9188,
      "step": 90
    },
    {
      "epoch": 0.020998477610373247,
      "grad_norm": 657.1996459960938,
      "learning_rate": 9.199999999999998e-06,
      "loss": 6.9113,
      "step": 100
    },
    {
      "epoch": 0.023098325371410572,
      "grad_norm": 1241.19482421875,
      "learning_rate": 1.02e-05,
      "loss": 6.8953,
      "step": 110
    },
    {
      "epoch": 0.025198173132447897,
      "grad_norm": 411.5585021972656,
      "learning_rate": 1.12e-05,
      "loss": 6.8474,
      "step": 120
    },
    {
      "epoch": 0.027298020893485223,
      "grad_norm": 1062.7774658203125,
      "learning_rate": 1.2199999999999998e-05,
      "loss": 6.8514,
      "step": 130
    },
    {
      "epoch": 0.029397868654522548,
      "grad_norm": 148.0970001220703,
      "learning_rate": 1.3199999999999997e-05,
      "loss": 6.8698,
      "step": 140
    },
    {
      "epoch": 0.03149771641555987,
      "grad_norm": 280.5868835449219,
      "learning_rate": 1.4199999999999998e-05,
      "loss": 6.8395,
      "step": 150
    },
    {
      "epoch": 0.0335975641765972,
      "grad_norm": 407.1022644042969,
      "learning_rate": 1.5199999999999998e-05,
      "loss": 6.8029,
      "step": 160
    },
    {
      "epoch": 0.035697411937634524,
      "grad_norm": 166.18153381347656,
      "learning_rate": 1.6199999999999997e-05,
      "loss": 6.7946,
      "step": 170
    },
    {
      "epoch": 0.03779725969867185,
      "grad_norm": 234.69805908203125,
      "learning_rate": 1.7199999999999998e-05,
      "loss": 6.7898,
      "step": 180
    },
    {
      "epoch": 0.03989710745970917,
      "grad_norm": 917.2261352539062,
      "learning_rate": 1.82e-05,
      "loss": 6.7996,
      "step": 190
    },
    {
      "epoch": 0.04199695522074649,
      "grad_norm": 237.35519409179688,
      "learning_rate": 1.92e-05,
      "loss": 6.7733,
      "step": 200
    },
    {
      "epoch": 0.04409680298178382,
      "grad_norm": 49.55766296386719,
      "learning_rate": 2.0199999999999996e-05,
      "loss": 6.7318,
      "step": 210
    },
    {
      "epoch": 0.046196650742821144,
      "grad_norm": 152.66554260253906,
      "learning_rate": 2.12e-05,
      "loss": 6.6867,
      "step": 220
    },
    {
      "epoch": 0.04829649850385847,
      "grad_norm": 284.306640625,
      "learning_rate": 2.2199999999999998e-05,
      "loss": 6.6555,
      "step": 230
    },
    {
      "epoch": 0.050396346264895794,
      "grad_norm": 127.93234252929688,
      "learning_rate": 2.3199999999999998e-05,
      "loss": 6.6263,
      "step": 240
    },
    {
      "epoch": 0.05249619402593312,
      "grad_norm": 119.77033996582031,
      "learning_rate": 2.42e-05,
      "loss": 6.649,
      "step": 250
    },
    {
      "epoch": 0.054596041786970445,
      "grad_norm": 389.56536865234375,
      "learning_rate": 2.52e-05,
      "loss": 6.6383,
      "step": 260
    },
    {
      "epoch": 0.05669588954800777,
      "grad_norm": 427.01580810546875,
      "learning_rate": 2.6199999999999996e-05,
      "loss": 6.593,
      "step": 270
    },
    {
      "epoch": 0.058795737309045096,
      "grad_norm": 222.83114624023438,
      "learning_rate": 2.72e-05,
      "loss": 6.5701,
      "step": 280
    },
    {
      "epoch": 0.06089558507008242,
      "grad_norm": 92.09854125976562,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 6.5495,
      "step": 290
    },
    {
      "epoch": 0.06299543283111975,
      "grad_norm": 137.1016082763672,
      "learning_rate": 2.9199999999999995e-05,
      "loss": 6.5433,
      "step": 300
    },
    {
      "epoch": 0.06509528059215706,
      "grad_norm": 165.11099243164062,
      "learning_rate": 3.02e-05,
      "loss": 6.5528,
      "step": 310
    },
    {
      "epoch": 0.0671951283531944,
      "grad_norm": 193.72972106933594,
      "learning_rate": 3.119999999999999e-05,
      "loss": 6.4983,
      "step": 320
    },
    {
      "epoch": 0.06929497611423172,
      "grad_norm": 168.2730255126953,
      "learning_rate": 3.22e-05,
      "loss": 6.4875,
      "step": 330
    },
    {
      "epoch": 0.07139482387526905,
      "grad_norm": 204.45751953125,
      "learning_rate": 3.3199999999999994e-05,
      "loss": 6.4667,
      "step": 340
    },
    {
      "epoch": 0.07349467163630637,
      "grad_norm": 200.1185302734375,
      "learning_rate": 3.42e-05,
      "loss": 6.4163,
      "step": 350
    },
    {
      "epoch": 0.0755945193973437,
      "grad_norm": 212.8289337158203,
      "learning_rate": 3.5199999999999995e-05,
      "loss": 6.4009,
      "step": 360
    },
    {
      "epoch": 0.07769436715838102,
      "grad_norm": 67.6531982421875,
      "learning_rate": 3.62e-05,
      "loss": 6.3942,
      "step": 370
    },
    {
      "epoch": 0.07979421491941834,
      "grad_norm": 103.08153533935547,
      "learning_rate": 3.7199999999999996e-05,
      "loss": 6.37,
      "step": 380
    },
    {
      "epoch": 0.08189406268045567,
      "grad_norm": 155.30517578125,
      "learning_rate": 3.8199999999999993e-05,
      "loss": 6.3541,
      "step": 390
    },
    {
      "epoch": 0.08399391044149299,
      "grad_norm": 78.574462890625,
      "learning_rate": 3.919999999999999e-05,
      "loss": 6.331,
      "step": 400
    },
    {
      "epoch": 0.08609375820253032,
      "grad_norm": 293.90155029296875,
      "learning_rate": 4.02e-05,
      "loss": 6.2978,
      "step": 410
    },
    {
      "epoch": 0.08819360596356764,
      "grad_norm": 133.43336486816406,
      "learning_rate": 4.12e-05,
      "loss": 6.2921,
      "step": 420
    },
    {
      "epoch": 0.09029345372460497,
      "grad_norm": 105.68527221679688,
      "learning_rate": 4.2199999999999996e-05,
      "loss": 6.2955,
      "step": 430
    },
    {
      "epoch": 0.09239330148564229,
      "grad_norm": 91.55309295654297,
      "learning_rate": 4.319999999999999e-05,
      "loss": 6.2592,
      "step": 440
    },
    {
      "epoch": 0.09449314924667962,
      "grad_norm": 72.15143585205078,
      "learning_rate": 4.42e-05,
      "loss": 6.2545,
      "step": 450
    },
    {
      "epoch": 0.09659299700771694,
      "grad_norm": 109.66728210449219,
      "learning_rate": 4.5199999999999994e-05,
      "loss": 6.2325,
      "step": 460
    },
    {
      "epoch": 0.09869284476875427,
      "grad_norm": 69.46297454833984,
      "learning_rate": 4.62e-05,
      "loss": 6.1691,
      "step": 470
    },
    {
      "epoch": 0.10079269252979159,
      "grad_norm": 137.57276916503906,
      "learning_rate": 4.7199999999999995e-05,
      "loss": 6.1748,
      "step": 480
    },
    {
      "epoch": 0.10289254029082892,
      "grad_norm": 142.85269165039062,
      "learning_rate": 4.82e-05,
      "loss": 6.1763,
      "step": 490
    },
    {
      "epoch": 0.10499238805186624,
      "grad_norm": 68.21001434326172,
      "learning_rate": 4.9199999999999997e-05,
      "loss": 6.1683,
      "step": 500
    },
    {
      "epoch": 0.10709223581290356,
      "grad_norm": 190.91734313964844,
      "learning_rate": 5.0199999999999994e-05,
      "loss": 6.1276,
      "step": 510
    },
    {
      "epoch": 0.10919208357394089,
      "grad_norm": 96.1802978515625,
      "learning_rate": 5.119999999999999e-05,
      "loss": 6.1462,
      "step": 520
    },
    {
      "epoch": 0.11129193133497821,
      "grad_norm": 67.83528137207031,
      "learning_rate": 5.2199999999999995e-05,
      "loss": 6.1067,
      "step": 530
    },
    {
      "epoch": 0.11339177909601554,
      "grad_norm": 42.651100158691406,
      "learning_rate": 5.32e-05,
      "loss": 6.0776,
      "step": 540
    },
    {
      "epoch": 0.11549162685705286,
      "grad_norm": 79.75711822509766,
      "learning_rate": 5.4199999999999996e-05,
      "loss": 6.0815,
      "step": 550
    },
    {
      "epoch": 0.11759147461809019,
      "grad_norm": 46.156211853027344,
      "learning_rate": 5.519999999999999e-05,
      "loss": 6.053,
      "step": 560
    },
    {
      "epoch": 0.11969132237912751,
      "grad_norm": 45.88727951049805,
      "learning_rate": 5.619999999999999e-05,
      "loss": 6.0348,
      "step": 570
    },
    {
      "epoch": 0.12179117014016484,
      "grad_norm": 46.44386672973633,
      "learning_rate": 5.72e-05,
      "loss": 6.0195,
      "step": 580
    },
    {
      "epoch": 0.12389101790120216,
      "grad_norm": 70.52837371826172,
      "learning_rate": 5.82e-05,
      "loss": 5.9979,
      "step": 590
    },
    {
      "epoch": 0.1259908656622395,
      "grad_norm": 59.506553649902344,
      "learning_rate": 5.9199999999999996e-05,
      "loss": 5.983,
      "step": 600
    },
    {
      "epoch": 0.12809071342327683,
      "grad_norm": 38.77995681762695,
      "learning_rate": 6.019999999999999e-05,
      "loss": 5.9576,
      "step": 610
    },
    {
      "epoch": 0.13019056118431413,
      "grad_norm": 62.45197677612305,
      "learning_rate": 6.12e-05,
      "loss": 5.951,
      "step": 620
    },
    {
      "epoch": 0.13229040894535146,
      "grad_norm": 46.94855499267578,
      "learning_rate": 6.22e-05,
      "loss": 5.9626,
      "step": 630
    },
    {
      "epoch": 0.1343902567063888,
      "grad_norm": 54.95368194580078,
      "learning_rate": 6.319999999999999e-05,
      "loss": 5.9268,
      "step": 640
    },
    {
      "epoch": 0.1364901044674261,
      "grad_norm": 47.36136245727539,
      "learning_rate": 6.419999999999999e-05,
      "loss": 5.8941,
      "step": 650
    },
    {
      "epoch": 0.13858995222846343,
      "grad_norm": 40.393096923828125,
      "learning_rate": 6.519999999999999e-05,
      "loss": 5.8787,
      "step": 660
    },
    {
      "epoch": 0.14068979998950076,
      "grad_norm": 41.42562484741211,
      "learning_rate": 6.62e-05,
      "loss": 5.8632,
      "step": 670
    },
    {
      "epoch": 0.1427896477505381,
      "grad_norm": 41.52008056640625,
      "learning_rate": 6.72e-05,
      "loss": 5.8166,
      "step": 680
    },
    {
      "epoch": 0.1448894955115754,
      "grad_norm": 60.34001922607422,
      "learning_rate": 6.819999999999999e-05,
      "loss": 5.8412,
      "step": 690
    },
    {
      "epoch": 0.14698934327261273,
      "grad_norm": 45.22390365600586,
      "learning_rate": 6.919999999999999e-05,
      "loss": 5.7904,
      "step": 700
    },
    {
      "epoch": 0.14908919103365006,
      "grad_norm": 40.19221496582031,
      "learning_rate": 7.02e-05,
      "loss": 5.7796,
      "step": 710
    },
    {
      "epoch": 0.1511890387946874,
      "grad_norm": 38.9415283203125,
      "learning_rate": 7.12e-05,
      "loss": 5.7776,
      "step": 720
    },
    {
      "epoch": 0.1532888865557247,
      "grad_norm": 36.409732818603516,
      "learning_rate": 7.219999999999999e-05,
      "loss": 5.7442,
      "step": 730
    },
    {
      "epoch": 0.15538873431676203,
      "grad_norm": 33.082157135009766,
      "learning_rate": 7.319999999999999e-05,
      "loss": 5.7215,
      "step": 740
    },
    {
      "epoch": 0.15748858207779937,
      "grad_norm": 40.2638053894043,
      "learning_rate": 7.419999999999999e-05,
      "loss": 5.6798,
      "step": 750
    },
    {
      "epoch": 0.15958842983883667,
      "grad_norm": 36.72783660888672,
      "learning_rate": 7.519999999999998e-05,
      "loss": 5.6984,
      "step": 760
    },
    {
      "epoch": 0.161688277599874,
      "grad_norm": 40.32900619506836,
      "learning_rate": 7.62e-05,
      "loss": 5.6429,
      "step": 770
    },
    {
      "epoch": 0.16378812536091134,
      "grad_norm": 31.345502853393555,
      "learning_rate": 7.72e-05,
      "loss": 5.6839,
      "step": 780
    },
    {
      "epoch": 0.16588797312194867,
      "grad_norm": 42.13299560546875,
      "learning_rate": 7.819999999999999e-05,
      "loss": 5.641,
      "step": 790
    },
    {
      "epoch": 0.16798782088298597,
      "grad_norm": 42.9160270690918,
      "learning_rate": 7.92e-05,
      "loss": 5.6259,
      "step": 800
    },
    {
      "epoch": 0.1700876686440233,
      "grad_norm": 33.14772415161133,
      "learning_rate": 8.019999999999998e-05,
      "loss": 5.6016,
      "step": 810
    },
    {
      "epoch": 0.17218751640506064,
      "grad_norm": 39.43216323852539,
      "learning_rate": 8.12e-05,
      "loss": 5.5918,
      "step": 820
    },
    {
      "epoch": 0.17428736416609797,
      "grad_norm": 31.36737823486328,
      "learning_rate": 8.22e-05,
      "loss": 5.5341,
      "step": 830
    },
    {
      "epoch": 0.17638721192713527,
      "grad_norm": 44.43039321899414,
      "learning_rate": 8.319999999999999e-05,
      "loss": 5.5423,
      "step": 840
    },
    {
      "epoch": 0.1784870596881726,
      "grad_norm": 38.169437408447266,
      "learning_rate": 8.42e-05,
      "loss": 5.5218,
      "step": 850
    },
    {
      "epoch": 0.18058690744920994,
      "grad_norm": 29.236454010009766,
      "learning_rate": 8.519999999999998e-05,
      "loss": 5.4835,
      "step": 860
    },
    {
      "epoch": 0.18268675521024727,
      "grad_norm": 35.421852111816406,
      "learning_rate": 8.62e-05,
      "loss": 5.4431,
      "step": 870
    },
    {
      "epoch": 0.18478660297128457,
      "grad_norm": 34.888694763183594,
      "learning_rate": 8.719999999999999e-05,
      "loss": 5.4332,
      "step": 880
    },
    {
      "epoch": 0.1868864507323219,
      "grad_norm": 30.417287826538086,
      "learning_rate": 8.819999999999999e-05,
      "loss": 5.418,
      "step": 890
    },
    {
      "epoch": 0.18898629849335924,
      "grad_norm": 32.38473892211914,
      "learning_rate": 8.92e-05,
      "loss": 5.4061,
      "step": 900
    }
  ],
  "logging_steps": 10,
  "max_steps": 100000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 21,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.87162480050176e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}