{
  "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": 159.0,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 24.9691,
      "step": 1
    },
    {
      "epoch": 0.91,
      "grad_norm": 70.5,
      "learning_rate": 7.692307692307693e-05,
      "loss": 23.7344,
      "step": 5
    },
    {
      "epoch": 0.91,
      "eval_loss": 7.958380222320557,
      "eval_runtime": 0.5623,
      "eval_samples_per_second": 3.557,
      "eval_steps_per_second": 1.778,
      "step": 5
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.75,
      "learning_rate": 0.00015384615384615385,
      "loss": 14.6026,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_loss": 6.82885217666626,
      "eval_runtime": 0.5598,
      "eval_samples_per_second": 3.573,
      "eval_steps_per_second": 1.786,
      "step": 11
    },
    {
      "epoch": 2.73,
      "grad_norm": 2.90625,
      "learning_rate": 0.00019984268150178167,
      "loss": 10.8118,
      "step": 15
    },
    {
      "epoch": 2.91,
      "eval_loss": 6.418484687805176,
      "eval_runtime": 0.5774,
      "eval_samples_per_second": 3.464,
      "eval_steps_per_second": 1.732,
      "step": 16
    },
    {
      "epoch": 3.64,
      "grad_norm": 4.8125,
      "learning_rate": 0.00019807852804032305,
      "loss": 10.8598,
      "step": 20
    },
    {
      "epoch": 4.0,
      "eval_loss": 5.106105327606201,
      "eval_runtime": 0.5656,
      "eval_samples_per_second": 3.536,
      "eval_steps_per_second": 1.768,
      "step": 22
    },
    {
      "epoch": 4.55,
      "grad_norm": 16.625,
      "learning_rate": 0.00019438833303083678,
      "loss": 7.9354,
      "step": 25
    },
    {
      "epoch": 4.91,
      "eval_loss": 1.7010552883148193,
      "eval_runtime": 0.5782,
      "eval_samples_per_second": 3.459,
      "eval_steps_per_second": 1.73,
      "step": 27
    },
    {
      "epoch": 5.45,
      "grad_norm": 2.21875,
      "learning_rate": 0.00018884456359788724,
      "loss": 2.0354,
      "step": 30
    },
    {
      "epoch": 6.0,
      "eval_loss": 1.4460557699203491,
      "eval_runtime": 0.5682,
      "eval_samples_per_second": 3.52,
      "eval_steps_per_second": 1.76,
      "step": 33
    },
    {
      "epoch": 6.36,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00018155608689592604,
      "loss": 1.4855,
      "step": 35
    },
    {
      "epoch": 6.91,
      "eval_loss": 1.3564677238464355,
      "eval_runtime": 0.5812,
      "eval_samples_per_second": 3.441,
      "eval_steps_per_second": 1.721,
      "step": 38
    },
    {
      "epoch": 7.27,
      "grad_norm": 1.203125,
      "learning_rate": 0.0001726660322034027,
      "loss": 1.326,
      "step": 40
    },
    {
      "epoch": 8.0,
      "eval_loss": 1.2935034036636353,
      "eval_runtime": 0.5674,
      "eval_samples_per_second": 3.525,
      "eval_steps_per_second": 1.762,
      "step": 44
    },
    {
      "epoch": 8.18,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00016234898018587337,
      "loss": 1.1375,
      "step": 45
    },
    {
      "epoch": 8.91,
      "eval_loss": 1.269553542137146,
      "eval_runtime": 0.5872,
      "eval_samples_per_second": 3.406,
      "eval_steps_per_second": 1.703,
      "step": 49
    },
    {
      "epoch": 9.09,
      "grad_norm": 1.0625,
      "learning_rate": 0.00015080753452465296,
      "loss": 1.0376,
      "step": 50
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000138268343236509,
      "loss": 0.9091,
      "step": 55
    },
    {
      "epoch": 10.0,
      "eval_loss": 1.2716257572174072,
      "eval_runtime": 0.5653,
      "eval_samples_per_second": 3.538,
      "eval_steps_per_second": 1.769,
      "step": 55
    },
    {
      "epoch": 10.91,
      "grad_norm": 0.59375,
      "learning_rate": 0.0001249776478167227,
      "loss": 0.8111,
      "step": 60
    },
    {
      "epoch": 10.91,
      "eval_loss": 1.2860848903656006,
      "eval_runtime": 0.5837,
      "eval_samples_per_second": 3.427,
      "eval_steps_per_second": 1.713,
      "step": 60
    },
    {
      "epoch": 11.82,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00011119644761033078,
      "loss": 0.689,
      "step": 65
    },
    {
      "epoch": 12.0,
      "eval_loss": 1.3148236274719238,
      "eval_runtime": 0.5682,
      "eval_samples_per_second": 3.52,
      "eval_steps_per_second": 1.76,
      "step": 66
    },
    {
      "epoch": 12.73,
      "grad_norm": 0.59375,
      "learning_rate": 9.719537437241312e-05,
      "loss": 0.6341,
      "step": 70
    },
    {
      "epoch": 12.91,
      "eval_loss": 1.3391039371490479,
      "eval_runtime": 0.7408,
      "eval_samples_per_second": 2.7,
      "eval_steps_per_second": 1.35,
      "step": 71
    },
    {
      "epoch": 13.64,
      "grad_norm": 0.8984375,
      "learning_rate": 8.324937766952638e-05,
      "loss": 0.5359,
      "step": 75
    },
    {
      "epoch": 14.0,
      "eval_loss": 1.4231812953948975,
      "eval_runtime": 0.5681,
      "eval_samples_per_second": 3.521,
      "eval_steps_per_second": 1.76,
      "step": 77
    },
    {
      "epoch": 14.55,
      "grad_norm": 1.0078125,
      "learning_rate": 6.963232548903853e-05,
      "loss": 0.4664,
      "step": 80
    },
    {
      "epoch": 14.91,
      "eval_loss": 1.510708212852478,
      "eval_runtime": 0.6308,
      "eval_samples_per_second": 3.171,
      "eval_steps_per_second": 1.585,
      "step": 82
    },
    {
      "epoch": 15.45,
      "grad_norm": 0.68359375,
      "learning_rate": 5.6611626088244194e-05,
      "loss": 0.3951,
      "step": 85
    },
    {
      "epoch": 16.0,
      "eval_loss": 1.6597082614898682,
      "eval_runtime": 0.5686,
      "eval_samples_per_second": 3.517,
      "eval_steps_per_second": 1.759,
      "step": 88
    },
    {
      "epoch": 16.36,
      "grad_norm": 0.67578125,
      "learning_rate": 4.444297669803981e-05,
      "loss": 0.3593,
      "step": 90
    },
    {
      "epoch": 16.91,
      "eval_loss": 1.9376537799835205,
      "eval_runtime": 0.5859,
      "eval_samples_per_second": 3.413,
      "eval_steps_per_second": 1.707,
      "step": 93
    },
    {
      "epoch": 17.27,
      "grad_norm": 0.53125,
      "learning_rate": 3.336534220479961e-05,
      "loss": 0.2802,
      "step": 95
    },
    {
      "epoch": 18.0,
      "eval_loss": 1.9024397134780884,
      "eval_runtime": 0.5686,
      "eval_samples_per_second": 3.518,
      "eval_steps_per_second": 1.759,
      "step": 99
    },
    {
      "epoch": 18.18,
      "grad_norm": 0.466796875,
      "learning_rate": 2.3596262417839255e-05,
      "loss": 0.2613,
      "step": 100
    },
    {
      "epoch": 18.91,
      "eval_loss": 2.098067283630371,
      "eval_runtime": 0.5848,
      "eval_samples_per_second": 3.42,
      "eval_steps_per_second": 1.71,
      "step": 104
    },
    {
      "epoch": 19.09,
      "grad_norm": 0.63671875,
      "learning_rate": 1.5327580077171587e-05,
      "loss": 0.2442,
      "step": 105
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.51171875,
      "learning_rate": 8.72167349386811e-06,
      "loss": 0.2262,
      "step": 110
    },
    {
      "epoch": 20.0,
      "eval_loss": 2.1472132205963135,
      "eval_runtime": 0.5636,
      "eval_samples_per_second": 3.548,
      "eval_steps_per_second": 1.774,
      "step": 110
    },
    {
      "epoch": 20.91,
      "grad_norm": 0.34375,
      "learning_rate": 3.908267805490051e-06,
      "loss": 0.2169,
      "step": 115
    },
    {
      "epoch": 20.91,
      "eval_loss": 2.1632509231567383,
      "eval_runtime": 0.5774,
      "eval_samples_per_second": 3.464,
      "eval_steps_per_second": 1.732,
      "step": 115
    },
    {
      "epoch": 21.82,
      "grad_norm": 0.333984375,
      "learning_rate": 9.818874663554357e-07,
      "loss": 0.2232,
      "step": 120
    },
    {
      "epoch": 22.0,
      "eval_loss": 2.159508466720581,
      "eval_runtime": 0.5656,
      "eval_samples_per_second": 3.536,
      "eval_steps_per_second": 1.768,
      "step": 121
    },
    {
      "epoch": 22.73,
      "grad_norm": 0.46875,
      "learning_rate": 0.0,
      "loss": 0.2096,
      "step": 125
    },
    {
      "epoch": 22.73,
      "eval_loss": 2.161546468734741,
      "eval_runtime": 0.563,
      "eval_samples_per_second": 3.553,
      "eval_steps_per_second": 1.776,
      "step": 125
    },
    {
      "epoch": 22.73,
      "step": 125,
      "total_flos": 1.929524923995259e+17,
      "train_loss": 3.26698664855957,
      "train_runtime": 470.6237,
      "train_samples_per_second": 4.675,
      "train_steps_per_second": 0.266
    }
  ],
  "logging_steps": 5,
  "max_steps": 125,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 25,
  "save_steps": 100,
  "total_flos": 1.929524923995259e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}