{
  "best_metric": 0.4463141858577728,
  "best_model_checkpoint": "mikhail-panzo/fil_b32_le3_s4000/checkpoint-500",
  "epoch": 77.77777777777777,
  "eval_steps": 500,
  "global_step": 3500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 3.548590898513794,
      "learning_rate": 2.5e-05,
      "loss": 0.7354,
      "step": 50
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 5.517350196838379,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 0.5892,
      "step": 100
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 4.143002033233643,
      "learning_rate": 7.45e-05,
      "loss": 0.532,
      "step": 150
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 6.414477825164795,
      "learning_rate": 9.95e-05,
      "loss": 0.5284,
      "step": 200
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 2.265143394470215,
      "learning_rate": 0.0001245,
      "loss": 0.517,
      "step": 250
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 4.954311847686768,
      "learning_rate": 0.0001495,
      "loss": 0.5295,
      "step": 300
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 2.5993571281433105,
      "learning_rate": 0.00017449999999999999,
      "loss": 0.517,
      "step": 350
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 3.27908992767334,
      "learning_rate": 0.00019950000000000002,
      "loss": 0.4939,
      "step": 400
    },
    {
      "epoch": 10.0,
      "grad_norm": 4.22811222076416,
      "learning_rate": 0.0002245,
      "loss": 0.5043,
      "step": 450
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 2.7051336765289307,
      "learning_rate": 0.0002495,
      "loss": 0.5028,
      "step": 500
    },
    {
      "epoch": 11.11111111111111,
      "eval_loss": 0.4463141858577728,
      "eval_runtime": 16.8864,
      "eval_samples_per_second": 9.416,
      "eval_steps_per_second": 1.184,
      "step": 500
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 3.0811643600463867,
      "learning_rate": 0.0002745,
      "loss": 0.545,
      "step": 550
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 7.707492828369141,
      "learning_rate": 0.0002995,
      "loss": 0.4979,
      "step": 600
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 5.85659122467041,
      "learning_rate": 0.00032450000000000003,
      "loss": 0.4916,
      "step": 650
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 8.300439834594727,
      "learning_rate": 0.0003495,
      "loss": 0.5112,
      "step": 700
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 10.664773941040039,
      "learning_rate": 0.0003745,
      "loss": 0.5312,
      "step": 750
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 6.4097161293029785,
      "learning_rate": 0.0003995,
      "loss": 0.6087,
      "step": 800
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 5.435739994049072,
      "learning_rate": 0.0004245,
      "loss": 0.6127,
      "step": 850
    },
    {
      "epoch": 20.0,
      "grad_norm": 7.4649834632873535,
      "learning_rate": 0.00044950000000000003,
      "loss": 0.6904,
      "step": 900
    },
    {
      "epoch": 21.11111111111111,
      "grad_norm": 10.81910228729248,
      "learning_rate": 0.0004745,
      "loss": 0.6948,
      "step": 950
    },
    {
      "epoch": 22.22222222222222,
      "grad_norm": 4.4550886154174805,
      "learning_rate": 0.0004995,
      "loss": 0.6348,
      "step": 1000
    },
    {
      "epoch": 22.22222222222222,
      "eval_loss": 0.5990382432937622,
      "eval_runtime": 9.0191,
      "eval_samples_per_second": 17.629,
      "eval_steps_per_second": 2.218,
      "step": 1000
    },
    {
      "epoch": 23.333333333333332,
      "grad_norm": 3.101961374282837,
      "learning_rate": 0.0005245,
      "loss": 0.6609,
      "step": 1050
    },
    {
      "epoch": 24.444444444444443,
      "grad_norm": 5.606886386871338,
      "learning_rate": 0.0005495,
      "loss": 0.6128,
      "step": 1100
    },
    {
      "epoch": 25.555555555555557,
      "grad_norm": 3.7152516841888428,
      "learning_rate": 0.0005745,
      "loss": 0.6197,
      "step": 1150
    },
    {
      "epoch": 26.666666666666668,
      "grad_norm": 2.9999840259552,
      "learning_rate": 0.0005995000000000001,
      "loss": 0.6346,
      "step": 1200
    },
    {
      "epoch": 27.77777777777778,
      "grad_norm": 3.5235953330993652,
      "learning_rate": 0.0006245000000000001,
      "loss": 0.6246,
      "step": 1250
    },
    {
      "epoch": 28.88888888888889,
      "grad_norm": 3.9665956497192383,
      "learning_rate": 0.0006495,
      "loss": 0.6143,
      "step": 1300
    },
    {
      "epoch": 30.0,
      "grad_norm": 2.920466661453247,
      "learning_rate": 0.000674,
      "loss": 0.6245,
      "step": 1350
    },
    {
      "epoch": 31.11111111111111,
      "grad_norm": 16.839616775512695,
      "learning_rate": 0.000699,
      "loss": 0.8179,
      "step": 1400
    },
    {
      "epoch": 32.22222222222222,
      "grad_norm": 0.23306678235530853,
      "learning_rate": 0.0007235000000000001,
      "loss": 1.8511,
      "step": 1450
    },
    {
      "epoch": 33.333333333333336,
      "grad_norm": 0.6476005911827087,
      "learning_rate": 0.000748,
      "loss": 1.7172,
      "step": 1500
    },
    {
      "epoch": 33.333333333333336,
      "eval_loss": 1.566503882408142,
      "eval_runtime": 8.82,
      "eval_samples_per_second": 18.027,
      "eval_steps_per_second": 2.268,
      "step": 1500
    },
    {
      "epoch": 34.44444444444444,
      "grad_norm": 0.22506840527057648,
      "learning_rate": 0.000773,
      "loss": 1.6751,
      "step": 1550
    },
    {
      "epoch": 35.55555555555556,
      "grad_norm": 0.2123740017414093,
      "learning_rate": 0.0007980000000000001,
      "loss": 1.585,
      "step": 1600
    },
    {
      "epoch": 36.666666666666664,
      "grad_norm": 0.2724120318889618,
      "learning_rate": 0.000823,
      "loss": 1.6075,
      "step": 1650
    },
    {
      "epoch": 37.77777777777778,
      "grad_norm": 0.1595434546470642,
      "learning_rate": 0.000848,
      "loss": 1.5939,
      "step": 1700
    },
    {
      "epoch": 38.888888888888886,
      "grad_norm": 0.329246461391449,
      "learning_rate": 0.000873,
      "loss": 1.6055,
      "step": 1750
    },
    {
      "epoch": 40.0,
      "grad_norm": 0.5109756588935852,
      "learning_rate": 0.000898,
      "loss": 1.5925,
      "step": 1800
    },
    {
      "epoch": 41.111111111111114,
      "grad_norm": 0.13621735572814941,
      "learning_rate": 0.0009230000000000001,
      "loss": 1.5823,
      "step": 1850
    },
    {
      "epoch": 42.22222222222222,
      "grad_norm": 0.13570882380008698,
      "learning_rate": 0.000948,
      "loss": 1.5806,
      "step": 1900
    },
    {
      "epoch": 43.333333333333336,
      "grad_norm": 0.10358592867851257,
      "learning_rate": 0.000973,
      "loss": 1.5762,
      "step": 1950
    },
    {
      "epoch": 44.44444444444444,
      "grad_norm": 0.2059299349784851,
      "learning_rate": 0.000998,
      "loss": 1.5863,
      "step": 2000
    },
    {
      "epoch": 44.44444444444444,
      "eval_loss": 1.5431957244873047,
      "eval_runtime": 8.5158,
      "eval_samples_per_second": 18.671,
      "eval_steps_per_second": 2.349,
      "step": 2000
    },
    {
      "epoch": 45.55555555555556,
      "grad_norm": 0.1365118771791458,
      "learning_rate": 0.000977,
      "loss": 1.578,
      "step": 2050
    },
    {
      "epoch": 46.666666666666664,
      "grad_norm": 0.482020765542984,
      "learning_rate": 0.0009519999999999999,
      "loss": 1.5794,
      "step": 2100
    },
    {
      "epoch": 47.77777777777778,
      "grad_norm": 0.0648043230175972,
      "learning_rate": 0.0009270000000000001,
      "loss": 1.6363,
      "step": 2150
    },
    {
      "epoch": 48.888888888888886,
      "grad_norm": 0.11886011809110641,
      "learning_rate": 0.000902,
      "loss": 1.5797,
      "step": 2200
    },
    {
      "epoch": 50.0,
      "grad_norm": 0.35273927450180054,
      "learning_rate": 0.0008770000000000001,
      "loss": 1.5799,
      "step": 2250
    },
    {
      "epoch": 51.111111111111114,
      "grad_norm": 0.3486323356628418,
      "learning_rate": 0.000852,
      "loss": 1.5786,
      "step": 2300
    },
    {
      "epoch": 52.22222222222222,
      "grad_norm": 0.10198960453271866,
      "learning_rate": 0.0008269999999999999,
      "loss": 1.5792,
      "step": 2350
    },
    {
      "epoch": 53.333333333333336,
      "grad_norm": 0.45619985461235046,
      "learning_rate": 0.0008020000000000001,
      "loss": 1.5791,
      "step": 2400
    },
    {
      "epoch": 54.44444444444444,
      "grad_norm": 0.18695257604122162,
      "learning_rate": 0.000777,
      "loss": 1.589,
      "step": 2450
    },
    {
      "epoch": 55.55555555555556,
      "grad_norm": 0.16741132736206055,
      "learning_rate": 0.0007520000000000001,
      "loss": 1.5854,
      "step": 2500
    },
    {
      "epoch": 55.55555555555556,
      "eval_loss": 1.5435634851455688,
      "eval_runtime": 8.0681,
      "eval_samples_per_second": 19.707,
      "eval_steps_per_second": 2.479,
      "step": 2500
    },
    {
      "epoch": 56.666666666666664,
      "grad_norm": 0.12719061970710754,
      "learning_rate": 0.000727,
      "loss": 1.5736,
      "step": 2550
    },
    {
      "epoch": 57.77777777777778,
      "grad_norm": 0.2572345435619354,
      "learning_rate": 0.0007019999999999999,
      "loss": 1.5808,
      "step": 2600
    },
    {
      "epoch": 58.888888888888886,
      "grad_norm": 0.13093958795070648,
      "learning_rate": 0.0006770000000000001,
      "loss": 1.5701,
      "step": 2650
    },
    {
      "epoch": 60.0,
      "grad_norm": 0.1683294177055359,
      "learning_rate": 0.000652,
      "loss": 1.6017,
      "step": 2700
    },
    {
      "epoch": 61.111111111111114,
      "grad_norm": 0.15231318771839142,
      "learning_rate": 0.0006270000000000001,
      "loss": 1.5809,
      "step": 2750
    },
    {
      "epoch": 62.22222222222222,
      "grad_norm": 0.1944093108177185,
      "learning_rate": 0.000602,
      "loss": 1.5782,
      "step": 2800
    },
    {
      "epoch": 63.333333333333336,
      "grad_norm": 0.3554971218109131,
      "learning_rate": 0.0005769999999999999,
      "loss": 1.5842,
      "step": 2850
    },
    {
      "epoch": 64.44444444444444,
      "grad_norm": 0.3123963475227356,
      "learning_rate": 0.0005520000000000001,
      "loss": 1.5645,
      "step": 2900
    },
    {
      "epoch": 65.55555555555556,
      "grad_norm": 0.14703120291233063,
      "learning_rate": 0.000527,
      "loss": 1.5861,
      "step": 2950
    },
    {
      "epoch": 66.66666666666667,
      "grad_norm": 0.363651305437088,
      "learning_rate": 0.0005020000000000001,
      "loss": 1.5647,
      "step": 3000
    },
    {
      "epoch": 66.66666666666667,
      "eval_loss": 1.5431376695632935,
      "eval_runtime": 7.9782,
      "eval_samples_per_second": 19.929,
      "eval_steps_per_second": 2.507,
      "step": 3000
    },
    {
      "epoch": 67.77777777777777,
      "grad_norm": 0.08024784922599792,
      "learning_rate": 0.000477,
      "loss": 1.5842,
      "step": 3050
    },
    {
      "epoch": 68.88888888888889,
      "grad_norm": 0.18207858502864838,
      "learning_rate": 0.00045200000000000004,
      "loss": 1.5793,
      "step": 3100
    },
    {
      "epoch": 70.0,
      "grad_norm": 0.17078548669815063,
      "learning_rate": 0.000427,
      "loss": 1.5718,
      "step": 3150
    },
    {
      "epoch": 71.11111111111111,
      "grad_norm": 0.3073755204677582,
      "learning_rate": 0.000402,
      "loss": 1.5741,
      "step": 3200
    },
    {
      "epoch": 72.22222222222223,
      "grad_norm": 0.06820886582136154,
      "learning_rate": 0.000377,
      "loss": 1.5748,
      "step": 3250
    },
    {
      "epoch": 73.33333333333333,
      "grad_norm": 0.034480493515729904,
      "learning_rate": 0.000352,
      "loss": 1.5807,
      "step": 3300
    },
    {
      "epoch": 74.44444444444444,
      "grad_norm": 0.06758826971054077,
      "learning_rate": 0.00032700000000000003,
      "loss": 1.5719,
      "step": 3350
    },
    {
      "epoch": 75.55555555555556,
      "grad_norm": 0.11566317081451416,
      "learning_rate": 0.000302,
      "loss": 1.5747,
      "step": 3400
    },
    {
      "epoch": 76.66666666666667,
      "grad_norm": 0.15969228744506836,
      "learning_rate": 0.000277,
      "loss": 1.5846,
      "step": 3450
    },
    {
      "epoch": 77.77777777777777,
      "grad_norm": 0.07962628453969955,
      "learning_rate": 0.000252,
      "loss": 1.5731,
      "step": 3500
    },
    {
      "epoch": 77.77777777777777,
      "eval_loss": 1.5472484827041626,
      "eval_runtime": 8.7234,
      "eval_samples_per_second": 18.227,
      "eval_steps_per_second": 2.293,
      "step": 3500
    }
  ],
  "logging_steps": 50,
  "max_steps": 4000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 89,
  "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": 2.408072794054013e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}