{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7171029042667623,
  "eval_steps": 2000,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04,
      "grad_norm": 2.6033709049224854,
      "learning_rate": 9.9e-07,
      "loss": 1.0789,
      "step": 100
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.7718708515167236,
      "learning_rate": 9.9e-07,
      "loss": 0.8366,
      "step": 200
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.209937572479248,
      "learning_rate": 9.8e-07,
      "loss": 0.6863,
      "step": 300
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.8079888820648193,
      "learning_rate": 9.698989898989898e-07,
      "loss": 0.6659,
      "step": 400
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.12679386138916,
      "learning_rate": 9.598989898989899e-07,
      "loss": 0.6546,
      "step": 500
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.909221887588501,
      "learning_rate": 9.497979797979798e-07,
      "loss": 0.6581,
      "step": 600
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.234121322631836,
      "learning_rate": 9.396969696969696e-07,
      "loss": 0.6224,
      "step": 700
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.9791226387023926,
      "learning_rate": 9.295959595959596e-07,
      "loss": 0.6139,
      "step": 800
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.557487726211548,
      "learning_rate": 9.194949494949495e-07,
      "loss": 0.6272,
      "step": 900
    },
    {
      "epoch": 0.36,
      "grad_norm": 3.946579694747925,
      "learning_rate": 9.093939393939394e-07,
      "loss": 0.6312,
      "step": 1000
    },
    {
      "epoch": 0.39,
      "grad_norm": 3.4085192680358887,
      "learning_rate": 8.992929292929292e-07,
      "loss": 0.6139,
      "step": 1100
    },
    {
      "epoch": 0.43,
      "grad_norm": 3.036348342895508,
      "learning_rate": 8.891919191919191e-07,
      "loss": 0.637,
      "step": 1200
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.595344066619873,
      "learning_rate": 8.790909090909091e-07,
      "loss": 0.6447,
      "step": 1300
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.76248025894165,
      "learning_rate": 8.68989898989899e-07,
      "loss": 0.6419,
      "step": 1400
    },
    {
      "epoch": 0.54,
      "grad_norm": 2.9645845890045166,
      "learning_rate": 8.588888888888888e-07,
      "loss": 0.6569,
      "step": 1500
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.357501268386841,
      "learning_rate": 8.487878787878787e-07,
      "loss": 0.6345,
      "step": 1600
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.0495171546936035,
      "learning_rate": 8.386868686868687e-07,
      "loss": 0.6004,
      "step": 1700
    },
    {
      "epoch": 0.65,
      "grad_norm": 3.0601418018341064,
      "learning_rate": 8.285858585858585e-07,
      "loss": 0.6444,
      "step": 1800
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.655466079711914,
      "learning_rate": 8.184848484848484e-07,
      "loss": 0.621,
      "step": 1900
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.7475666999816895,
      "learning_rate": 8.083838383838384e-07,
      "loss": 0.6554,
      "step": 2000
    },
    {
      "epoch": 0.72,
      "eval_loss": 0.7412441372871399,
      "eval_runtime": 565.2186,
      "eval_samples_per_second": 1.769,
      "eval_steps_per_second": 0.442,
      "step": 2000
    }
  ],
  "logging_steps": 100,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 2000,
  "total_flos": 1.88564197343232e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}