{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 22.727272727272727,
  "eval_steps": 500,
  "global_step": 125,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.18,
      "grad_norm": 220.0,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 24.9691,
      "step": 1
    },
    {
      "epoch": 0.91,
      "grad_norm": 90.0,
      "learning_rate": 7.692307692307693e-05,
      "loss": 23.6212,
      "step": 5
    },
    {
      "epoch": 0.91,
      "eval_loss": 8.00199031829834,
      "eval_runtime": 0.5608,
      "eval_samples_per_second": 3.566,
      "eval_steps_per_second": 1.783,
      "step": 5
    },
    {
      "epoch": 1.82,
      "grad_norm": 14.8125,
      "learning_rate": 0.00015384615384615385,
      "loss": 14.6688,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_loss": 6.809861183166504,
      "eval_runtime": 0.5624,
      "eval_samples_per_second": 3.556,
      "eval_steps_per_second": 1.778,
      "step": 11
    },
    {
      "epoch": 2.73,
      "grad_norm": 5.625,
      "learning_rate": 0.00019984268150178167,
      "loss": 10.8277,
      "step": 15
    },
    {
      "epoch": 2.91,
      "eval_loss": 6.458514213562012,
      "eval_runtime": 0.5773,
      "eval_samples_per_second": 3.465,
      "eval_steps_per_second": 1.732,
      "step": 16
    },
    {
      "epoch": 3.64,
      "grad_norm": 7.65625,
      "learning_rate": 0.00019807852804032305,
      "loss": 10.965,
      "step": 20
    },
    {
      "epoch": 4.0,
      "eval_loss": 5.275852203369141,
      "eval_runtime": 0.5667,
      "eval_samples_per_second": 3.529,
      "eval_steps_per_second": 1.765,
      "step": 22
    },
    {
      "epoch": 4.55,
      "grad_norm": 32.75,
      "learning_rate": 0.00019438833303083678,
      "loss": 8.3233,
      "step": 25
    },
    {
      "epoch": 4.91,
      "eval_loss": 1.6938855648040771,
      "eval_runtime": 0.5862,
      "eval_samples_per_second": 3.412,
      "eval_steps_per_second": 1.706,
      "step": 27
    },
    {
      "epoch": 5.45,
      "grad_norm": 2.890625,
      "learning_rate": 0.00018884456359788724,
      "loss": 2.2795,
      "step": 30
    },
    {
      "epoch": 6.0,
      "eval_loss": 1.4539892673492432,
      "eval_runtime": 0.5675,
      "eval_samples_per_second": 3.524,
      "eval_steps_per_second": 1.762,
      "step": 33
    },
    {
      "epoch": 6.36,
      "grad_norm": 35.25,
      "learning_rate": 0.00018155608689592604,
      "loss": 1.5047,
      "step": 35
    },
    {
      "epoch": 6.91,
      "eval_loss": 1.3612087965011597,
      "eval_runtime": 0.5882,
      "eval_samples_per_second": 3.4,
      "eval_steps_per_second": 1.7,
      "step": 38
    },
    {
      "epoch": 7.27,
      "grad_norm": 1.65625,
      "learning_rate": 0.0001726660322034027,
      "loss": 1.3243,
      "step": 40
    },
    {
      "epoch": 8.0,
      "eval_loss": 1.288640022277832,
      "eval_runtime": 0.5698,
      "eval_samples_per_second": 3.51,
      "eval_steps_per_second": 1.755,
      "step": 44
    },
    {
      "epoch": 8.18,
      "grad_norm": 1.015625,
      "learning_rate": 0.00016234898018587337,
      "loss": 1.1264,
      "step": 45
    },
    {
      "epoch": 8.91,
      "eval_loss": 1.2782788276672363,
      "eval_runtime": 0.59,
      "eval_samples_per_second": 3.39,
      "eval_steps_per_second": 1.695,
      "step": 49
    },
    {
      "epoch": 9.09,
      "grad_norm": 2.1875,
      "learning_rate": 0.00015080753452465296,
      "loss": 1.0387,
      "step": 50
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000138268343236509,
      "loss": 0.9122,
      "step": 55
    },
    {
      "epoch": 10.0,
      "eval_loss": 1.2739648818969727,
      "eval_runtime": 0.5659,
      "eval_samples_per_second": 3.534,
      "eval_steps_per_second": 1.767,
      "step": 55
    },
    {
      "epoch": 10.91,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0001249776478167227,
      "loss": 0.8184,
      "step": 60
    },
    {
      "epoch": 10.91,
      "eval_loss": 1.2853541374206543,
      "eval_runtime": 0.5892,
      "eval_samples_per_second": 3.395,
      "eval_steps_per_second": 1.697,
      "step": 60
    },
    {
      "epoch": 11.82,
      "grad_norm": 0.96875,
      "learning_rate": 0.00011119644761033078,
      "loss": 0.6918,
      "step": 65
    },
    {
      "epoch": 12.0,
      "eval_loss": 1.31352698802948,
      "eval_runtime": 0.5674,
      "eval_samples_per_second": 3.525,
      "eval_steps_per_second": 1.762,
      "step": 66
    },
    {
      "epoch": 12.73,
      "grad_norm": 0.92578125,
      "learning_rate": 9.719537437241312e-05,
      "loss": 0.6194,
      "step": 70
    },
    {
      "epoch": 12.91,
      "eval_loss": 1.343058705329895,
      "eval_runtime": 0.7364,
      "eval_samples_per_second": 2.716,
      "eval_steps_per_second": 1.358,
      "step": 71
    },
    {
      "epoch": 13.64,
      "grad_norm": 1.1875,
      "learning_rate": 8.324937766952638e-05,
      "loss": 0.5176,
      "step": 75
    },
    {
      "epoch": 14.0,
      "eval_loss": 1.4736580848693848,
      "eval_runtime": 0.5693,
      "eval_samples_per_second": 3.513,
      "eval_steps_per_second": 1.757,
      "step": 77
    },
    {
      "epoch": 14.55,
      "grad_norm": 0.9296875,
      "learning_rate": 6.963232548903853e-05,
      "loss": 0.4514,
      "step": 80
    },
    {
      "epoch": 14.91,
      "eval_loss": 1.7112184762954712,
      "eval_runtime": 0.6242,
      "eval_samples_per_second": 3.204,
      "eval_steps_per_second": 1.602,
      "step": 82
    },
    {
      "epoch": 15.45,
      "grad_norm": 1.34375,
      "learning_rate": 5.6611626088244194e-05,
      "loss": 0.3759,
      "step": 85
    },
    {
      "epoch": 16.0,
      "eval_loss": 1.8429330587387085,
      "eval_runtime": 0.5694,
      "eval_samples_per_second": 3.513,
      "eval_steps_per_second": 1.756,
      "step": 88
    },
    {
      "epoch": 16.36,
      "grad_norm": 0.84375,
      "learning_rate": 4.444297669803981e-05,
      "loss": 0.3464,
      "step": 90
    },
    {
      "epoch": 16.91,
      "eval_loss": 1.899384617805481,
      "eval_runtime": 0.5863,
      "eval_samples_per_second": 3.411,
      "eval_steps_per_second": 1.706,
      "step": 93
    },
    {
      "epoch": 17.27,
      "grad_norm": 0.8203125,
      "learning_rate": 3.336534220479961e-05,
      "loss": 0.2681,
      "step": 95
    },
    {
      "epoch": 18.0,
      "eval_loss": 1.9583137035369873,
      "eval_runtime": 0.5683,
      "eval_samples_per_second": 3.519,
      "eval_steps_per_second": 1.76,
      "step": 99
    },
    {
      "epoch": 18.18,
      "grad_norm": 0.73828125,
      "learning_rate": 2.3596262417839255e-05,
      "loss": 0.2487,
      "step": 100
    },
    {
      "epoch": 18.91,
      "eval_loss": 2.162316083908081,
      "eval_runtime": 0.5845,
      "eval_samples_per_second": 3.422,
      "eval_steps_per_second": 1.711,
      "step": 104
    },
    {
      "epoch": 19.09,
      "grad_norm": 0.875,
      "learning_rate": 1.5327580077171587e-05,
      "loss": 0.2322,
      "step": 105
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.6484375,
      "learning_rate": 8.72167349386811e-06,
      "loss": 0.2122,
      "step": 110
    },
    {
      "epoch": 20.0,
      "eval_loss": 2.213620901107788,
      "eval_runtime": 0.5651,
      "eval_samples_per_second": 3.539,
      "eval_steps_per_second": 1.769,
      "step": 110
    },
    {
      "epoch": 20.91,
      "grad_norm": 0.50390625,
      "learning_rate": 3.908267805490051e-06,
      "loss": 0.2036,
      "step": 115
    },
    {
      "epoch": 20.91,
      "eval_loss": 2.2149863243103027,
      "eval_runtime": 0.5812,
      "eval_samples_per_second": 3.441,
      "eval_steps_per_second": 1.721,
      "step": 115
    },
    {
      "epoch": 21.82,
      "grad_norm": 0.458984375,
      "learning_rate": 9.818874663554357e-07,
      "loss": 0.2098,
      "step": 120
    },
    {
      "epoch": 22.0,
      "eval_loss": 2.2188880443573,
      "eval_runtime": 0.5673,
      "eval_samples_per_second": 3.525,
      "eval_steps_per_second": 1.763,
      "step": 121
    },
    {
      "epoch": 22.73,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0,
      "loss": 0.1955,
      "step": 125
    },
    {
      "epoch": 22.73,
      "eval_loss": 2.209489345550537,
      "eval_runtime": 0.5657,
      "eval_samples_per_second": 3.536,
      "eval_steps_per_second": 1.768,
      "step": 125
    },
    {
      "epoch": 22.73,
      "step": 125,
      "total_flos": 1.917235948819579e+17,
      "train_loss": 3.2900945229530336,
      "train_runtime": 472.0874,
      "train_samples_per_second": 4.66,
      "train_steps_per_second": 0.265
    }
  ],
  "logging_steps": 5,
  "max_steps": 125,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 25,
  "save_steps": 100,
  "total_flos": 1.917235948819579e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}