{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0125,
      "grad_norm": 7352.2265625,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 175.5381,
      "step": 25
    },
    {
      "epoch": 0.025,
      "grad_norm": 2155.41015625,
      "learning_rate": 3.266666666666667e-05,
      "loss": 174.6277,
      "step": 50
    },
    {
      "epoch": 0.0375,
      "grad_norm": 29.340621948242188,
      "learning_rate": 4.933333333333334e-05,
      "loss": 157.9852,
      "step": 75
    },
    {
      "epoch": 0.05,
      "grad_norm": 11.753185272216797,
      "learning_rate": 6.6e-05,
      "loss": 133.1257,
      "step": 100
    },
    {
      "epoch": 0.05,
      "eval_accuracy": 0.08756986554511534,
      "eval_loss": 7.706512928009033,
      "eval_runtime": 7.3236,
      "eval_samples_per_second": 66.224,
      "eval_steps_per_second": 2.185,
      "step": 100
    },
    {
      "epoch": 0.0625,
      "grad_norm": 11.1610746383667,
      "learning_rate": 8.266666666666667e-05,
      "loss": 119.0136,
      "step": 125
    },
    {
      "epoch": 0.075,
      "grad_norm": 7.337241172790527,
      "learning_rate": 9.933333333333334e-05,
      "loss": 113.2304,
      "step": 150
    },
    {
      "epoch": 0.0875,
      "grad_norm": 9.895074844360352,
      "learning_rate": 9.995847987378953e-05,
      "loss": 109.5068,
      "step": 175
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.935585975646973,
      "learning_rate": 9.982700328363471e-05,
      "loss": 107.0186,
      "step": 200
    },
    {
      "epoch": 0.1,
      "eval_accuracy": 0.13644318107230674,
      "eval_loss": 6.609853267669678,
      "eval_runtime": 7.4218,
      "eval_samples_per_second": 65.348,
      "eval_steps_per_second": 2.156,
      "step": 200
    },
    {
      "epoch": 0.1125,
      "grad_norm": 17.207761764526367,
      "learning_rate": 9.96057350657239e-05,
      "loss": 105.3798,
      "step": 225
    },
    {
      "epoch": 0.125,
      "grad_norm": 37.589744567871094,
      "learning_rate": 9.929507396034141e-05,
      "loss": 103.9933,
      "step": 250
    },
    {
      "epoch": 0.1375,
      "grad_norm": 48.09913635253906,
      "learning_rate": 9.889557979979694e-05,
      "loss": 103.0857,
      "step": 275
    },
    {
      "epoch": 0.15,
      "grad_norm": 18.950061798095703,
      "learning_rate": 9.840797249956985e-05,
      "loss": 102.3462,
      "step": 300
    },
    {
      "epoch": 0.15,
      "eval_accuracy": 0.14476797317970908,
      "eval_loss": 6.367699146270752,
      "eval_runtime": 7.5111,
      "eval_samples_per_second": 64.571,
      "eval_steps_per_second": 2.13,
      "step": 300
    },
    {
      "epoch": 0.1625,
      "grad_norm": 54.86927032470703,
      "learning_rate": 9.783313076097283e-05,
      "loss": 101.6082,
      "step": 325
    },
    {
      "epoch": 0.175,
      "grad_norm": 14.090791702270508,
      "learning_rate": 9.717209048767277e-05,
      "loss": 100.8465,
      "step": 350
    },
    {
      "epoch": 0.1875,
      "grad_norm": 42.630802154541016,
      "learning_rate": 9.642604291892226e-05,
      "loss": 99.9092,
      "step": 375
    },
    {
      "epoch": 0.2,
      "grad_norm": 12.524935722351074,
      "learning_rate": 9.559633248286604e-05,
      "loss": 99.0524,
      "step": 400
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.15310791228876244,
      "eval_loss": 6.214518070220947,
      "eval_runtime": 7.3913,
      "eval_samples_per_second": 65.618,
      "eval_steps_per_second": 2.165,
      "step": 400
    },
    {
      "epoch": 0.2125,
      "grad_norm": 22.39153480529785,
      "learning_rate": 9.468445437379055e-05,
      "loss": 98.4639,
      "step": 425
    },
    {
      "epoch": 0.225,
      "grad_norm": 20.86402702331543,
      "learning_rate": 9.369205185768272e-05,
      "loss": 97.6668,
      "step": 450
    },
    {
      "epoch": 0.2375,
      "grad_norm": 15.403814315795898,
      "learning_rate": 9.262091331095375e-05,
      "loss": 97.1439,
      "step": 475
    },
    {
      "epoch": 0.25,
      "grad_norm": 18.481685638427734,
      "learning_rate": 9.147296899766376e-05,
      "loss": 96.6031,
      "step": 500
    },
    {
      "epoch": 0.25,
      "eval_accuracy": 0.1584204706678313,
      "eval_loss": 6.100996494293213,
      "eval_runtime": 7.4101,
      "eval_samples_per_second": 65.451,
      "eval_steps_per_second": 2.159,
      "step": 500
    },
    {
      "epoch": 0.2625,
      "grad_norm": 19.892127990722656,
      "learning_rate": 9.025028759105558e-05,
      "loss": 96.2492,
      "step": 525
    },
    {
      "epoch": 0.275,
      "grad_norm": 20.713096618652344,
      "learning_rate": 8.895507244566574e-05,
      "loss": 96.3436,
      "step": 550
    },
    {
      "epoch": 0.2875,
      "grad_norm": 12.466355323791504,
      "learning_rate": 8.758965762673064e-05,
      "loss": 96.5737,
      "step": 575
    },
    {
      "epoch": 0.3,
      "grad_norm": 21.678146362304688,
      "learning_rate": 8.615650370404327e-05,
      "loss": 95.9314,
      "step": 600
    },
    {
      "epoch": 0.3,
      "eval_accuracy": 0.163982449674087,
      "eval_loss": 5.986855983734131,
      "eval_runtime": 7.6235,
      "eval_samples_per_second": 63.619,
      "eval_steps_per_second": 2.099,
      "step": 600
    },
    {
      "epoch": 0.3125,
      "grad_norm": 20.58213233947754,
      "learning_rate": 8.465819331784009e-05,
      "loss": 96.2906,
      "step": 625
    },
    {
      "epoch": 0.325,
      "grad_norm": 14.315438270568848,
      "learning_rate": 8.30974265247088e-05,
      "loss": 95.9487,
      "step": 650
    },
    {
      "epoch": 0.3375,
      "grad_norm": 12.04114818572998,
      "learning_rate": 8.147701593190385e-05,
      "loss": 95.4309,
      "step": 675
    },
    {
      "epoch": 0.35,
      "grad_norm": 22.478317260742188,
      "learning_rate": 7.979988162883804e-05,
      "loss": 95.0101,
      "step": 700
    },
    {
      "epoch": 0.35,
      "eval_accuracy": 0.16826703154110645,
      "eval_loss": 5.88900899887085,
      "eval_runtime": 7.5208,
      "eval_samples_per_second": 64.487,
      "eval_steps_per_second": 2.127,
      "step": 700
    },
    {
      "epoch": 0.3625,
      "grad_norm": 50.81674575805664,
      "learning_rate": 7.806904592488408e-05,
      "loss": 94.8858,
      "step": 725
    },
    {
      "epoch": 0.375,
      "grad_norm": 20.65149688720703,
      "learning_rate": 7.628762790296879e-05,
      "loss": 95.0119,
      "step": 750
    },
    {
      "epoch": 0.3875,
      "grad_norm": 22.1734619140625,
      "learning_rate": 7.445883779877482e-05,
      "loss": 94.6604,
      "step": 775
    },
    {
      "epoch": 0.4,
      "grad_norm": 20.18880844116211,
      "learning_rate": 7.258597121567904e-05,
      "loss": 94.2643,
      "step": 800
    },
    {
      "epoch": 0.4,
      "eval_accuracy": 0.17154484269838785,
      "eval_loss": 5.828437328338623,
      "eval_runtime": 7.4208,
      "eval_samples_per_second": 65.357,
      "eval_steps_per_second": 2.156,
      "step": 800
    },
    {
      "epoch": 0.4125,
      "grad_norm": 35.1305046081543,
      "learning_rate": 7.067240318585242e-05,
      "loss": 94.3216,
      "step": 825
    },
    {
      "epoch": 0.425,
      "grad_norm": 9.143704414367676,
      "learning_rate": 6.872158208822398e-05,
      "loss": 93.9239,
      "step": 850
    },
    {
      "epoch": 0.4375,
      "grad_norm": 9.495540618896484,
      "learning_rate": 6.673702343426894e-05,
      "loss": 93.8027,
      "step": 875
    },
    {
      "epoch": 0.45,
      "grad_norm": 19.112733840942383,
      "learning_rate": 6.472230353281925e-05,
      "loss": 93.2266,
      "step": 900
    },
    {
      "epoch": 0.45,
      "eval_accuracy": 0.17456010582705153,
      "eval_loss": 5.77442741394043,
      "eval_runtime": 7.4706,
      "eval_samples_per_second": 64.921,
      "eval_steps_per_second": 2.142,
      "step": 900
    },
    {
      "epoch": 0.4625,
      "grad_norm": 13.373994827270508,
      "learning_rate": 6.268105304531354e-05,
      "loss": 92.7674,
      "step": 925
    },
    {
      "epoch": 0.475,
      "grad_norm": 20.701358795166016,
      "learning_rate": 6.061695044309973e-05,
      "loss": 92.4968,
      "step": 950
    },
    {
      "epoch": 0.4875,
      "grad_norm": 13.155655860900879,
      "learning_rate": 5.8533715378581e-05,
      "loss": 92.2234,
      "step": 975
    },
    {
      "epoch": 0.5,
      "grad_norm": 46.88108444213867,
      "learning_rate": 5.643510198215082e-05,
      "loss": 92.0125,
      "step": 1000
    },
    {
      "epoch": 0.5,
      "eval_accuracy": 0.17633533442055147,
      "eval_loss": 5.7336883544921875,
      "eval_runtime": 7.3739,
      "eval_samples_per_second": 65.773,
      "eval_steps_per_second": 2.17,
      "step": 1000
    },
    {
      "epoch": 0.5125,
      "grad_norm": 482.2004699707031,
      "learning_rate": 5.432489209699614e-05,
      "loss": 93.2148,
      "step": 1025
    },
    {
      "epoch": 0.525,
      "grad_norm": 297.8095703125,
      "learning_rate": 5.220688846396047e-05,
      "loss": 95.0004,
      "step": 1050
    },
    {
      "epoch": 0.5375,
      "grad_norm": 332.8581848144531,
      "learning_rate": 5.0084907868747746e-05,
      "loss": 95.3143,
      "step": 1075
    },
    {
      "epoch": 0.55,
      "grad_norm": 9733.95703125,
      "learning_rate": 4.796277426381657e-05,
      "loss": 95.8613,
      "step": 1100
    },
    {
      "epoch": 0.55,
      "eval_accuracy": 0.1554284329416992,
      "eval_loss": 6.006833553314209,
      "eval_runtime": 7.2985,
      "eval_samples_per_second": 66.452,
      "eval_steps_per_second": 2.192,
      "step": 1100
    },
    {
      "epoch": 0.5625,
      "grad_norm": 1343.75537109375,
      "learning_rate": 4.5844311877359384e-05,
      "loss": 96.7118,
      "step": 1125
    },
    {
      "epoch": 0.575,
      "grad_norm": 234.01663208007812,
      "learning_rate": 4.373333832178478e-05,
      "loss": 97.3733,
      "step": 1150
    },
    {
      "epoch": 0.5875,
      "grad_norm": 868.69921875,
      "learning_rate": 4.1633657714122e-05,
      "loss": 97.8172,
      "step": 1175
    },
    {
      "epoch": 0.6,
      "grad_norm": 370.631591796875,
      "learning_rate": 3.954905382074478e-05,
      "loss": 98.1012,
      "step": 1200
    },
    {
      "epoch": 0.6,
      "eval_accuracy": 0.14907376084903992,
      "eval_loss": 6.105465888977051,
      "eval_runtime": 7.4782,
      "eval_samples_per_second": 64.855,
      "eval_steps_per_second": 2.14,
      "step": 1200
    },
    {
      "epoch": 0.6125,
      "grad_norm": 843.0389404296875,
      "learning_rate": 3.748328323876868e-05,
      "loss": 97.997,
      "step": 1225
    },
    {
      "epoch": 0.625,
      "grad_norm": 657.6774291992188,
      "learning_rate": 3.544006862640916e-05,
      "loss": 97.6659,
      "step": 1250
    },
    {
      "epoch": 0.6375,
      "grad_norm": 736.9610595703125,
      "learning_rate": 3.342309199449991e-05,
      "loss": 97.2111,
      "step": 1275
    },
    {
      "epoch": 0.65,
      "grad_norm": 629.177490234375,
      "learning_rate": 3.143598807126035e-05,
      "loss": 97.152,
      "step": 1300
    },
    {
      "epoch": 0.65,
      "eval_accuracy": 0.15122059588304496,
      "eval_loss": 6.054599761962891,
      "eval_runtime": 7.3568,
      "eval_samples_per_second": 65.926,
      "eval_steps_per_second": 2.175,
      "step": 1300
    },
    {
      "epoch": 0.6625,
      "grad_norm": 690.4661254882812,
      "learning_rate": 2.9482337752269752e-05,
      "loss": 97.7027,
      "step": 1325
    },
    {
      "epoch": 0.675,
      "grad_norm": 853.2005615234375,
      "learning_rate": 2.7565661647451425e-05,
      "loss": 98.0302,
      "step": 1350
    },
    {
      "epoch": 0.6875,
      "grad_norm": 652.2321166992188,
      "learning_rate": 2.568941373669562e-05,
      "loss": 97.5237,
      "step": 1375
    },
    {
      "epoch": 0.7,
      "grad_norm": 17031.59375,
      "learning_rate": 2.3856975145554318e-05,
      "loss": 97.8734,
      "step": 1400
    },
    {
      "epoch": 0.7,
      "eval_accuracy": 0.14743788467072944,
      "eval_loss": 6.094700336456299,
      "eval_runtime": 7.6558,
      "eval_samples_per_second": 63.351,
      "eval_steps_per_second": 2.09,
      "step": 1400
    },
    {
      "epoch": 0.7125,
      "grad_norm": 795.8380126953125,
      "learning_rate": 2.2071648052224408e-05,
      "loss": 98.055,
      "step": 1425
    },
    {
      "epoch": 0.725,
      "grad_norm": 1413.1805419921875,
      "learning_rate": 2.0336649736799495e-05,
      "loss": 98.1603,
      "step": 1450
    },
    {
      "epoch": 0.7375,
      "grad_norm": 746.7064208984375,
      "learning_rate": 1.865510678351361e-05,
      "loss": 97.9418,
      "step": 1475
    },
    {
      "epoch": 0.75,
      "grad_norm": 505.42279052734375,
      "learning_rate": 1.7030049446425132e-05,
      "loss": 97.4964,
      "step": 1500
    },
    {
      "epoch": 0.75,
      "eval_accuracy": 0.15097016545574804,
      "eval_loss": 6.041877746582031,
      "eval_runtime": 7.513,
      "eval_samples_per_second": 64.555,
      "eval_steps_per_second": 2.13,
      "step": 1500
    },
    {
      "epoch": 0.7625,
      "grad_norm": 2972.94580078125,
      "learning_rate": 1.5464406188694174e-05,
      "loss": 97.7231,
      "step": 1525
    },
    {
      "epoch": 0.775,
      "grad_norm": 2009.724609375,
      "learning_rate": 1.3960998405293963e-05,
      "loss": 97.6005,
      "step": 1550
    },
    {
      "epoch": 0.7875,
      "grad_norm": 955.8128051757812,
      "learning_rate": 1.2522535338666486e-05,
      "loss": 97.5096,
      "step": 1575
    },
    {
      "epoch": 0.8,
      "grad_norm": 1708.3966064453125,
      "learning_rate": 1.1151609196484447e-05,
      "loss": 97.562,
      "step": 1600
    },
    {
      "epoch": 0.8,
      "eval_accuracy": 0.15179921134611404,
      "eval_loss": 6.027195930480957,
      "eval_runtime": 7.3648,
      "eval_samples_per_second": 65.853,
      "eval_steps_per_second": 2.172,
      "step": 1600
    },
    {
      "epoch": 0.8125,
      "grad_norm": 249.3505096435547,
      "learning_rate": 9.850690480317837e-06,
      "loss": 97.4989,
      "step": 1625
    },
    {
      "epoch": 0.825,
      "grad_norm": 493.7890930175781,
      "learning_rate": 8.622123533623077e-06,
      "loss": 97.3748,
      "step": 1650
    },
    {
      "epoch": 0.8375,
      "grad_norm": 1148.812744140625,
      "learning_rate": 7.468122317077786e-06,
      "loss": 97.3686,
      "step": 1675
    },
    {
      "epoch": 0.85,
      "grad_norm": 719.5657958984375,
      "learning_rate": 6.390766418873812e-06,
      "loss": 97.3206,
      "step": 1700
    },
    {
      "epoch": 0.85,
      "eval_accuracy": 0.15286253086201587,
      "eval_loss": 6.017110824584961,
      "eval_runtime": 7.5198,
      "eval_samples_per_second": 64.496,
      "eval_steps_per_second": 2.128,
      "step": 1700
    },
    {
      "epoch": 0.8625,
      "grad_norm": 585.3778686523438,
      "learning_rate": 5.391997307158919e-06,
      "loss": 97.2298,
      "step": 1725
    },
    {
      "epoch": 0.875,
      "grad_norm": 2988.042236328125,
      "learning_rate": 4.473614831379824e-06,
      "loss": 97.108,
      "step": 1750
    },
    {
      "epoch": 0.8875,
      "grad_norm": 3200.808837890625,
      "learning_rate": 3.6372739788319843e-06,
      "loss": 97.0287,
      "step": 1775
    },
    {
      "epoch": 0.9,
      "grad_norm": 263.6023254394531,
      "learning_rate": 2.88448189226046e-06,
      "loss": 96.8684,
      "step": 1800
    },
    {
      "epoch": 0.9,
      "eval_accuracy": 0.15323817650296123,
      "eval_loss": 6.016338348388672,
      "eval_runtime": 14.8413,
      "eval_samples_per_second": 32.679,
      "eval_steps_per_second": 1.078,
      "step": 1800
    },
    {
      "epoch": 0.9125,
      "grad_norm": 300.458984375,
      "learning_rate": 2.216595153886969e-06,
      "loss": 96.5413,
      "step": 1825
    },
    {
      "epoch": 0.925,
      "grad_norm": 1235.066162109375,
      "learning_rate": 1.6348173407569557e-06,
      "loss": 96.6509,
      "step": 1850
    },
    {
      "epoch": 0.9375,
      "grad_norm": 1469.02978515625,
      "learning_rate": 1.1401968558123976e-06,
      "loss": 96.5367,
      "step": 1875
    },
    {
      "epoch": 0.95,
      "grad_norm": 632.1422119140625,
      "learning_rate": 7.336250385988896e-07,
      "loss": 96.472,
      "step": 1900
    },
    {
      "epoch": 0.95,
      "eval_accuracy": 0.15278578605365067,
      "eval_loss": 6.0192670822143555,
      "eval_runtime": 7.5713,
      "eval_samples_per_second": 64.058,
      "eval_steps_per_second": 2.113,
      "step": 1900
    },
    {
      "epoch": 0.9625,
      "grad_norm": 2200.51904296875,
      "learning_rate": 4.1583455901149647e-07,
      "loss": 96.5353,
      "step": 1925
    },
    {
      "epoch": 0.975,
      "grad_norm": 1819.2476806640625,
      "learning_rate": 1.8739809697409517e-07,
      "loss": 96.0391,
      "step": 1950
    },
    {
      "epoch": 0.9875,
      "grad_norm": 1034.934326171875,
      "learning_rate": 4.872731043143453e-08,
      "loss": 96.1378,
      "step": 1975
    },
    {
      "epoch": 1.0,
      "grad_norm": 477.2402648925781,
      "learning_rate": 7.209351372550188e-11,
      "loss": 96.5051,
      "step": 2000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.15278376645343053,
      "eval_loss": 6.019127368927002,
      "eval_runtime": 7.2999,
      "eval_samples_per_second": 66.44,
      "eval_steps_per_second": 2.192,
      "step": 2000
    }
  ],
  "logging_steps": 25,
  "max_steps": 2000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 300,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.337781380972544e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}