{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.414368184733804,
  "global_step": 40000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08,
      "learning_rate": 4.959910198845414e-05,
      "loss": 13.3635,
      "step": 500
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.919820397690827e-05,
      "loss": 3.169,
      "step": 1000
    },
    {
      "epoch": 0.16,
      "eval_loss": 0.662490725517273,
      "eval_runtime": 80.6872,
      "eval_samples_per_second": 62.922,
      "eval_steps_per_second": 31.467,
      "step": 1000
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.8797305965362415e-05,
      "loss": 0.578,
      "step": 1500
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.839640795381655e-05,
      "loss": 0.2929,
      "step": 2000
    },
    {
      "epoch": 0.32,
      "eval_loss": 0.18057043850421906,
      "eval_runtime": 80.3642,
      "eval_samples_per_second": 63.175,
      "eval_steps_per_second": 31.594,
      "step": 2000
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.799550994227069e-05,
      "loss": 0.2481,
      "step": 2500
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.759461193072483e-05,
      "loss": 0.2159,
      "step": 3000
    },
    {
      "epoch": 0.48,
      "eval_loss": 0.1447182148694992,
      "eval_runtime": 80.0634,
      "eval_samples_per_second": 63.412,
      "eval_steps_per_second": 31.712,
      "step": 3000
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.719371391917896e-05,
      "loss": 0.2131,
      "step": 3500
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.6792815907633104e-05,
      "loss": 0.1955,
      "step": 4000
    },
    {
      "epoch": 0.64,
      "eval_loss": 0.13617677986621857,
      "eval_runtime": 80.8812,
      "eval_samples_per_second": 62.771,
      "eval_steps_per_second": 31.392,
      "step": 4000
    },
    {
      "epoch": 0.72,
      "learning_rate": 4.639191789608724e-05,
      "loss": 0.1799,
      "step": 4500
    },
    {
      "epoch": 0.8,
      "learning_rate": 4.5991019884541375e-05,
      "loss": 0.1785,
      "step": 5000
    },
    {
      "epoch": 0.8,
      "eval_loss": 0.12113407254219055,
      "eval_runtime": 81.0258,
      "eval_samples_per_second": 62.659,
      "eval_steps_per_second": 31.336,
      "step": 5000
    },
    {
      "epoch": 0.88,
      "learning_rate": 4.559012187299551e-05,
      "loss": 0.1664,
      "step": 5500
    },
    {
      "epoch": 0.96,
      "learning_rate": 4.5189223861449645e-05,
      "loss": 0.1661,
      "step": 6000
    },
    {
      "epoch": 0.96,
      "eval_loss": 0.1217775046825409,
      "eval_runtime": 80.4348,
      "eval_samples_per_second": 63.119,
      "eval_steps_per_second": 31.566,
      "step": 6000
    },
    {
      "epoch": 1.04,
      "learning_rate": 4.478832584990379e-05,
      "loss": 0.152,
      "step": 6500
    },
    {
      "epoch": 1.12,
      "learning_rate": 4.438742783835793e-05,
      "loss": 0.1429,
      "step": 7000
    },
    {
      "epoch": 1.12,
      "eval_loss": 0.11830633133649826,
      "eval_runtime": 80.5584,
      "eval_samples_per_second": 63.023,
      "eval_steps_per_second": 31.518,
      "step": 7000
    },
    {
      "epoch": 1.2,
      "learning_rate": 4.3986529826812064e-05,
      "loss": 0.1436,
      "step": 7500
    },
    {
      "epoch": 1.28,
      "learning_rate": 4.35856318152662e-05,
      "loss": 0.138,
      "step": 8000
    },
    {
      "epoch": 1.28,
      "eval_loss": 0.11041316390037537,
      "eval_runtime": 80.3972,
      "eval_samples_per_second": 63.149,
      "eval_steps_per_second": 31.581,
      "step": 8000
    },
    {
      "epoch": 1.36,
      "learning_rate": 4.3184733803720335e-05,
      "loss": 0.139,
      "step": 8500
    },
    {
      "epoch": 1.44,
      "learning_rate": 4.278383579217447e-05,
      "loss": 0.1328,
      "step": 9000
    },
    {
      "epoch": 1.44,
      "eval_loss": 0.11453410983085632,
      "eval_runtime": 80.9914,
      "eval_samples_per_second": 62.686,
      "eval_steps_per_second": 31.349,
      "step": 9000
    },
    {
      "epoch": 1.52,
      "learning_rate": 4.238293778062861e-05,
      "loss": 0.1336,
      "step": 9500
    },
    {
      "epoch": 1.6,
      "learning_rate": 4.198203976908275e-05,
      "loss": 0.1257,
      "step": 10000
    },
    {
      "epoch": 1.6,
      "eval_loss": 0.11141891777515411,
      "eval_runtime": 80.764,
      "eval_samples_per_second": 62.862,
      "eval_steps_per_second": 31.437,
      "step": 10000
    },
    {
      "epoch": 1.68,
      "learning_rate": 4.158114175753688e-05,
      "loss": 0.129,
      "step": 10500
    },
    {
      "epoch": 1.76,
      "learning_rate": 4.118024374599102e-05,
      "loss": 0.1283,
      "step": 11000
    },
    {
      "epoch": 1.76,
      "eval_loss": 0.10540181398391724,
      "eval_runtime": 80.8099,
      "eval_samples_per_second": 62.826,
      "eval_steps_per_second": 31.419,
      "step": 11000
    },
    {
      "epoch": 1.84,
      "learning_rate": 4.077934573444516e-05,
      "loss": 0.1329,
      "step": 11500
    },
    {
      "epoch": 1.92,
      "learning_rate": 4.03784477228993e-05,
      "loss": 0.1233,
      "step": 12000
    },
    {
      "epoch": 1.92,
      "eval_loss": 0.10773167759180069,
      "eval_runtime": 80.5271,
      "eval_samples_per_second": 63.047,
      "eval_steps_per_second": 31.53,
      "step": 12000
    },
    {
      "epoch": 2.0,
      "learning_rate": 3.9977549711353436e-05,
      "loss": 0.121,
      "step": 12500
    },
    {
      "epoch": 2.08,
      "learning_rate": 3.957665169980757e-05,
      "loss": 0.1022,
      "step": 13000
    },
    {
      "epoch": 2.08,
      "eval_loss": 0.1054486483335495,
      "eval_runtime": 80.727,
      "eval_samples_per_second": 62.891,
      "eval_steps_per_second": 31.452,
      "step": 13000
    },
    {
      "epoch": 2.16,
      "learning_rate": 3.917575368826171e-05,
      "loss": 0.113,
      "step": 13500
    },
    {
      "epoch": 2.25,
      "learning_rate": 3.877485567671584e-05,
      "loss": 0.1015,
      "step": 14000
    },
    {
      "epoch": 2.25,
      "eval_loss": 0.10687392950057983,
      "eval_runtime": 80.3483,
      "eval_samples_per_second": 63.187,
      "eval_steps_per_second": 31.6,
      "step": 14000
    },
    {
      "epoch": 2.33,
      "learning_rate": 3.8373957665169984e-05,
      "loss": 0.1052,
      "step": 14500
    },
    {
      "epoch": 2.41,
      "learning_rate": 3.797305965362412e-05,
      "loss": 0.1013,
      "step": 15000
    },
    {
      "epoch": 2.41,
      "eval_loss": 0.0990850031375885,
      "eval_runtime": 80.592,
      "eval_samples_per_second": 62.996,
      "eval_steps_per_second": 31.504,
      "step": 15000
    },
    {
      "epoch": 2.49,
      "learning_rate": 3.7572161642078254e-05,
      "loss": 0.1107,
      "step": 15500
    },
    {
      "epoch": 2.57,
      "learning_rate": 3.7171263630532396e-05,
      "loss": 0.1024,
      "step": 16000
    },
    {
      "epoch": 2.57,
      "eval_loss": 0.10092990845441818,
      "eval_runtime": 81.1078,
      "eval_samples_per_second": 62.596,
      "eval_steps_per_second": 31.304,
      "step": 16000
    },
    {
      "epoch": 2.65,
      "learning_rate": 3.677036561898653e-05,
      "loss": 0.0986,
      "step": 16500
    },
    {
      "epoch": 2.73,
      "learning_rate": 3.6369467607440674e-05,
      "loss": 0.1011,
      "step": 17000
    },
    {
      "epoch": 2.73,
      "eval_loss": 0.09937073290348053,
      "eval_runtime": 79.8753,
      "eval_samples_per_second": 63.562,
      "eval_steps_per_second": 31.787,
      "step": 17000
    },
    {
      "epoch": 2.81,
      "learning_rate": 3.596856959589481e-05,
      "loss": 0.1009,
      "step": 17500
    },
    {
      "epoch": 2.89,
      "learning_rate": 3.5567671584348944e-05,
      "loss": 0.1011,
      "step": 18000
    },
    {
      "epoch": 2.89,
      "eval_loss": 0.09842894226312637,
      "eval_runtime": 80.4877,
      "eval_samples_per_second": 63.078,
      "eval_steps_per_second": 31.545,
      "step": 18000
    },
    {
      "epoch": 2.97,
      "learning_rate": 3.516677357280308e-05,
      "loss": 0.1008,
      "step": 18500
    },
    {
      "epoch": 3.05,
      "learning_rate": 3.4765875561257214e-05,
      "loss": 0.0935,
      "step": 19000
    },
    {
      "epoch": 3.05,
      "eval_loss": 0.1000189483165741,
      "eval_runtime": 80.5915,
      "eval_samples_per_second": 62.997,
      "eval_steps_per_second": 31.505,
      "step": 19000
    },
    {
      "epoch": 3.13,
      "learning_rate": 3.436497754971135e-05,
      "loss": 0.0847,
      "step": 19500
    },
    {
      "epoch": 3.21,
      "learning_rate": 3.396407953816549e-05,
      "loss": 0.086,
      "step": 20000
    },
    {
      "epoch": 3.21,
      "eval_loss": 0.10353543609380722,
      "eval_runtime": 80.3489,
      "eval_samples_per_second": 63.187,
      "eval_steps_per_second": 31.6,
      "step": 20000
    },
    {
      "epoch": 3.29,
      "learning_rate": 3.3563181526619633e-05,
      "loss": 0.0858,
      "step": 20500
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.316228351507377e-05,
      "loss": 0.0868,
      "step": 21000
    },
    {
      "epoch": 3.37,
      "eval_loss": 0.10058429092168808,
      "eval_runtime": 80.7731,
      "eval_samples_per_second": 62.855,
      "eval_steps_per_second": 31.434,
      "step": 21000
    },
    {
      "epoch": 3.45,
      "learning_rate": 3.2761385503527904e-05,
      "loss": 0.0822,
      "step": 21500
    },
    {
      "epoch": 3.53,
      "learning_rate": 3.236048749198204e-05,
      "loss": 0.0854,
      "step": 22000
    },
    {
      "epoch": 3.53,
      "eval_loss": 0.09908822923898697,
      "eval_runtime": 80.4062,
      "eval_samples_per_second": 63.142,
      "eval_steps_per_second": 31.577,
      "step": 22000
    },
    {
      "epoch": 3.61,
      "learning_rate": 3.195958948043618e-05,
      "loss": 0.0835,
      "step": 22500
    },
    {
      "epoch": 3.69,
      "learning_rate": 3.1558691468890316e-05,
      "loss": 0.0898,
      "step": 23000
    },
    {
      "epoch": 3.69,
      "eval_loss": 0.09145330637693405,
      "eval_runtime": 81.1442,
      "eval_samples_per_second": 62.568,
      "eval_steps_per_second": 31.29,
      "step": 23000
    },
    {
      "epoch": 3.77,
      "learning_rate": 3.115779345734445e-05,
      "loss": 0.0876,
      "step": 23500
    },
    {
      "epoch": 3.85,
      "learning_rate": 3.0756895445798587e-05,
      "loss": 0.08,
      "step": 24000
    },
    {
      "epoch": 3.85,
      "eval_loss": 0.09377244859933853,
      "eval_runtime": 80.6911,
      "eval_samples_per_second": 62.919,
      "eval_steps_per_second": 31.466,
      "step": 24000
    },
    {
      "epoch": 3.93,
      "learning_rate": 3.0355997434252725e-05,
      "loss": 0.091,
      "step": 24500
    },
    {
      "epoch": 4.01,
      "learning_rate": 2.9955099422706867e-05,
      "loss": 0.0778,
      "step": 25000
    },
    {
      "epoch": 4.01,
      "eval_loss": 0.09943201392889023,
      "eval_runtime": 80.4527,
      "eval_samples_per_second": 63.105,
      "eval_steps_per_second": 31.559,
      "step": 25000
    },
    {
      "epoch": 4.09,
      "learning_rate": 2.9554201411161002e-05,
      "loss": 0.0704,
      "step": 25500
    },
    {
      "epoch": 4.17,
      "learning_rate": 2.915330339961514e-05,
      "loss": 0.0699,
      "step": 26000
    },
    {
      "epoch": 4.17,
      "eval_loss": 0.09294962137937546,
      "eval_runtime": 80.2201,
      "eval_samples_per_second": 63.288,
      "eval_steps_per_second": 31.65,
      "step": 26000
    },
    {
      "epoch": 4.25,
      "learning_rate": 2.8752405388069276e-05,
      "loss": 0.079,
      "step": 26500
    },
    {
      "epoch": 4.33,
      "learning_rate": 2.835150737652341e-05,
      "loss": 0.0754,
      "step": 27000
    },
    {
      "epoch": 4.33,
      "eval_loss": 0.09366082400083542,
      "eval_runtime": 80.2272,
      "eval_samples_per_second": 63.283,
      "eval_steps_per_second": 31.648,
      "step": 27000
    },
    {
      "epoch": 4.41,
      "learning_rate": 2.7950609364977553e-05,
      "loss": 0.0707,
      "step": 27500
    },
    {
      "epoch": 4.49,
      "learning_rate": 2.754971135343169e-05,
      "loss": 0.0727,
      "step": 28000
    },
    {
      "epoch": 4.49,
      "eval_loss": 0.09205422550439835,
      "eval_runtime": 80.3418,
      "eval_samples_per_second": 63.193,
      "eval_steps_per_second": 31.602,
      "step": 28000
    },
    {
      "epoch": 4.57,
      "learning_rate": 2.7148813341885827e-05,
      "loss": 0.0723,
      "step": 28500
    },
    {
      "epoch": 4.65,
      "learning_rate": 2.6747915330339962e-05,
      "loss": 0.0718,
      "step": 29000
    },
    {
      "epoch": 4.65,
      "eval_loss": 0.09299702942371368,
      "eval_runtime": 80.7661,
      "eval_samples_per_second": 62.861,
      "eval_steps_per_second": 31.436,
      "step": 29000
    },
    {
      "epoch": 4.73,
      "learning_rate": 2.6347017318794097e-05,
      "loss": 0.0725,
      "step": 29500
    },
    {
      "epoch": 4.81,
      "learning_rate": 2.594611930724824e-05,
      "loss": 0.0741,
      "step": 30000
    },
    {
      "epoch": 4.81,
      "eval_loss": 0.09144437313079834,
      "eval_runtime": 80.7469,
      "eval_samples_per_second": 62.875,
      "eval_steps_per_second": 31.444,
      "step": 30000
    },
    {
      "epoch": 4.89,
      "learning_rate": 2.5545221295702375e-05,
      "loss": 0.0708,
      "step": 30500
    },
    {
      "epoch": 4.97,
      "learning_rate": 2.5144323284156513e-05,
      "loss": 0.0687,
      "step": 31000
    },
    {
      "epoch": 4.97,
      "eval_loss": 0.0920889675617218,
      "eval_runtime": 80.5972,
      "eval_samples_per_second": 62.992,
      "eval_steps_per_second": 31.502,
      "step": 31000
    },
    {
      "epoch": 5.05,
      "learning_rate": 2.474342527261065e-05,
      "loss": 0.0677,
      "step": 31500
    },
    {
      "epoch": 5.13,
      "learning_rate": 2.4342527261064787e-05,
      "loss": 0.0617,
      "step": 32000
    },
    {
      "epoch": 5.13,
      "eval_loss": 0.09358620643615723,
      "eval_runtime": 80.47,
      "eval_samples_per_second": 63.092,
      "eval_steps_per_second": 31.552,
      "step": 32000
    },
    {
      "epoch": 5.21,
      "learning_rate": 2.3941629249518922e-05,
      "loss": 0.0591,
      "step": 32500
    },
    {
      "epoch": 5.29,
      "learning_rate": 2.354073123797306e-05,
      "loss": 0.0609,
      "step": 33000
    },
    {
      "epoch": 5.29,
      "eval_loss": 0.09032619744539261,
      "eval_runtime": 80.985,
      "eval_samples_per_second": 62.691,
      "eval_steps_per_second": 31.351,
      "step": 33000
    },
    {
      "epoch": 5.37,
      "learning_rate": 2.31398332264272e-05,
      "loss": 0.0657,
      "step": 33500
    },
    {
      "epoch": 5.45,
      "learning_rate": 2.2738935214881335e-05,
      "loss": 0.0652,
      "step": 34000
    },
    {
      "epoch": 5.45,
      "eval_loss": 0.0909246951341629,
      "eval_runtime": 81.1024,
      "eval_samples_per_second": 62.6,
      "eval_steps_per_second": 31.306,
      "step": 34000
    },
    {
      "epoch": 5.53,
      "learning_rate": 2.233803720333547e-05,
      "loss": 0.0629,
      "step": 34500
    },
    {
      "epoch": 5.61,
      "learning_rate": 2.1937139191789612e-05,
      "loss": 0.0641,
      "step": 35000
    },
    {
      "epoch": 5.61,
      "eval_loss": 0.09035832434892654,
      "eval_runtime": 80.259,
      "eval_samples_per_second": 63.258,
      "eval_steps_per_second": 31.635,
      "step": 35000
    },
    {
      "epoch": 5.69,
      "learning_rate": 2.1536241180243747e-05,
      "loss": 0.0642,
      "step": 35500
    },
    {
      "epoch": 5.77,
      "learning_rate": 2.1135343168697885e-05,
      "loss": 0.06,
      "step": 36000
    },
    {
      "epoch": 5.77,
      "eval_loss": 0.09021405130624771,
      "eval_runtime": 80.1923,
      "eval_samples_per_second": 63.31,
      "eval_steps_per_second": 31.661,
      "step": 36000
    },
    {
      "epoch": 5.85,
      "learning_rate": 2.073444515715202e-05,
      "loss": 0.063,
      "step": 36500
    },
    {
      "epoch": 5.93,
      "learning_rate": 2.033354714560616e-05,
      "loss": 0.0643,
      "step": 37000
    },
    {
      "epoch": 5.93,
      "eval_loss": 0.0934990718960762,
      "eval_runtime": 79.9682,
      "eval_samples_per_second": 63.488,
      "eval_steps_per_second": 31.75,
      "step": 37000
    },
    {
      "epoch": 6.01,
      "learning_rate": 1.9932649134060298e-05,
      "loss": 0.0641,
      "step": 37500
    },
    {
      "epoch": 6.09,
      "learning_rate": 1.9531751122514433e-05,
      "loss": 0.0552,
      "step": 38000
    },
    {
      "epoch": 6.09,
      "eval_loss": 0.09127607196569443,
      "eval_runtime": 81.2206,
      "eval_samples_per_second": 62.509,
      "eval_steps_per_second": 31.261,
      "step": 38000
    },
    {
      "epoch": 6.17,
      "learning_rate": 1.913085311096857e-05,
      "loss": 0.0586,
      "step": 38500
    },
    {
      "epoch": 6.25,
      "learning_rate": 1.8729955099422707e-05,
      "loss": 0.0517,
      "step": 39000
    },
    {
      "epoch": 6.25,
      "eval_loss": 0.09465406835079193,
      "eval_runtime": 80.2376,
      "eval_samples_per_second": 63.275,
      "eval_steps_per_second": 31.644,
      "step": 39000
    },
    {
      "epoch": 6.33,
      "learning_rate": 1.8329057087876845e-05,
      "loss": 0.055,
      "step": 39500
    },
    {
      "epoch": 6.41,
      "learning_rate": 1.7928159076330984e-05,
      "loss": 0.0592,
      "step": 40000
    },
    {
      "epoch": 6.41,
      "eval_loss": 0.08911187946796417,
      "eval_runtime": 80.0512,
      "eval_samples_per_second": 63.422,
      "eval_steps_per_second": 31.717,
      "step": 40000
    }
  ],
  "max_steps": 62360,
  "num_train_epochs": 10,
  "total_flos": 1.9182220110515405e+17,
  "trial_name": null,
  "trial_params": null
}