{
  "best_metric": 1.6614694595336914,
  "best_model_checkpoint": "outputs/checkpoint-348",
  "epoch": 9.978494623655914,
  "eval_steps": 500,
  "global_step": 348,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5734767025089605,
      "grad_norm": 1.2473632097244263,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.9249,
      "step": 20
    },
    {
      "epoch": 0.974910394265233,
      "eval_loss": 2.817430019378662,
      "eval_runtime": 11.5884,
      "eval_samples_per_second": 32.101,
      "eval_steps_per_second": 4.056,
      "step": 34
    },
    {
      "epoch": 1.146953405017921,
      "grad_norm": 1.330991268157959,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.8932,
      "step": 40
    },
    {
      "epoch": 1.7204301075268817,
      "grad_norm": 1.5549895763397217,
      "learning_rate": 6e-06,
      "loss": 2.8287,
      "step": 60
    },
    {
      "epoch": 1.978494623655914,
      "eval_loss": 2.5992982387542725,
      "eval_runtime": 11.5924,
      "eval_samples_per_second": 32.09,
      "eval_steps_per_second": 4.054,
      "step": 69
    },
    {
      "epoch": 2.293906810035842,
      "grad_norm": 1.9673638343811035,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.6659,
      "step": 80
    },
    {
      "epoch": 2.867383512544803,
      "grad_norm": 1.759926199913025,
      "learning_rate": 1e-05,
      "loss": 2.3145,
      "step": 100
    },
    {
      "epoch": 2.982078853046595,
      "eval_loss": 2.0121684074401855,
      "eval_runtime": 11.5907,
      "eval_samples_per_second": 32.095,
      "eval_steps_per_second": 4.055,
      "step": 104
    },
    {
      "epoch": 3.4408602150537635,
      "grad_norm": 0.7806220650672913,
      "learning_rate": 9.896320793787106e-06,
      "loss": 2.0158,
      "step": 120
    },
    {
      "epoch": 3.985663082437276,
      "eval_loss": 1.8575435876846313,
      "eval_runtime": 11.5927,
      "eval_samples_per_second": 32.089,
      "eval_steps_per_second": 4.054,
      "step": 139
    },
    {
      "epoch": 4.014336917562724,
      "grad_norm": 0.6744422316551208,
      "learning_rate": 9.589582926268798e-06,
      "loss": 1.9297,
      "step": 140
    },
    {
      "epoch": 4.587813620071684,
      "grad_norm": 0.636353075504303,
      "learning_rate": 9.092507332892968e-06,
      "loss": 1.8752,
      "step": 160
    },
    {
      "epoch": 4.989247311827957,
      "eval_loss": 1.7877445220947266,
      "eval_runtime": 11.5902,
      "eval_samples_per_second": 32.096,
      "eval_steps_per_second": 4.055,
      "step": 174
    },
    {
      "epoch": 5.161290322580645,
      "grad_norm": 0.6430570483207703,
      "learning_rate": 8.425708574839221e-06,
      "loss": 1.813,
      "step": 180
    },
    {
      "epoch": 5.734767025089606,
      "grad_norm": 0.6457915306091309,
      "learning_rate": 7.616839918483061e-06,
      "loss": 1.7878,
      "step": 200
    },
    {
      "epoch": 5.992831541218638,
      "eval_loss": 1.7403738498687744,
      "eval_runtime": 11.5876,
      "eval_samples_per_second": 32.103,
      "eval_steps_per_second": 4.056,
      "step": 209
    },
    {
      "epoch": 6.308243727598566,
      "grad_norm": 0.6910040974617004,
      "learning_rate": 6.699446507913083e-06,
      "loss": 1.7346,
      "step": 220
    },
    {
      "epoch": 6.881720430107527,
      "grad_norm": 0.673575758934021,
      "learning_rate": 5.711574191366427e-06,
      "loss": 1.7293,
      "step": 240
    },
    {
      "epoch": 6.996415770609319,
      "eval_loss": 1.705881118774414,
      "eval_runtime": 11.588,
      "eval_samples_per_second": 32.102,
      "eval_steps_per_second": 4.056,
      "step": 244
    },
    {
      "epoch": 7.455197132616488,
      "grad_norm": 0.6435667276382446,
      "learning_rate": 4.694191695890788e-06,
      "loss": 1.7172,
      "step": 260
    },
    {
      "epoch": 8.0,
      "eval_loss": 1.6822781562805176,
      "eval_runtime": 11.5832,
      "eval_samples_per_second": 32.115,
      "eval_steps_per_second": 4.058,
      "step": 279
    },
    {
      "epoch": 8.028673835125447,
      "grad_norm": 0.5970215201377869,
      "learning_rate": 3.689491585304491e-06,
      "loss": 1.6713,
      "step": 280
    },
    {
      "epoch": 8.602150537634408,
      "grad_norm": 0.734722375869751,
      "learning_rate": 2.7391404635865725e-06,
      "loss": 1.6777,
      "step": 300
    },
    {
      "epoch": 8.974910394265233,
      "eval_loss": 1.6685205698013306,
      "eval_runtime": 11.5892,
      "eval_samples_per_second": 32.099,
      "eval_steps_per_second": 4.056,
      "step": 313
    },
    {
      "epoch": 9.175627240143369,
      "grad_norm": 0.6750782132148743,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 1.6782,
      "step": 320
    },
    {
      "epoch": 9.74910394265233,
      "grad_norm": 0.8320801854133606,
      "learning_rate": 1.1552473733031893e-06,
      "loss": 1.6629,
      "step": 340
    },
    {
      "epoch": 9.978494623655914,
      "eval_loss": 1.6614694595336914,
      "eval_runtime": 11.5856,
      "eval_samples_per_second": 32.109,
      "eval_steps_per_second": 4.057,
      "step": 348
    }
  ],
  "logging_steps": 20,
  "max_steps": 408,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 12,
  "save_steps": 500,
  "total_flos": 1.835118238875648e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}