trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 22,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09,
      "learning_rate": 0.0,
      "loss": 6.3097,
      "step": 1
    },
    {
      "epoch": 0.18,
      "learning_rate": 2e-05,
      "loss": 0.6861,
      "step": 2
    },
    {
      "epoch": 0.27,
      "learning_rate": 2e-05,
      "loss": 0.7619,
      "step": 3
    },
    {
      "epoch": 0.36,
      "learning_rate": 2e-05,
      "loss": 0.4096,
      "step": 4
    },
    {
      "epoch": 0.45,
      "learning_rate": 2e-05,
      "loss": 0.8462,
      "step": 5
    },
    {
      "epoch": 0.55,
      "learning_rate": 2e-05,
      "loss": 0.383,
      "step": 6
    },
    {
      "epoch": 0.64,
      "learning_rate": 2e-05,
      "loss": 0.49,
      "step": 7
    },
    {
      "epoch": 0.73,
      "learning_rate": 2e-05,
      "loss": 0.4643,
      "step": 8
    },
    {
      "epoch": 0.82,
      "learning_rate": 2e-05,
      "loss": 0.3448,
      "step": 9
    },
    {
      "epoch": 0.91,
      "learning_rate": 2e-05,
      "loss": 0.3445,
      "step": 10
    },
    {
      "epoch": 1.0,
      "learning_rate": 2e-05,
      "loss": 0.3333,
      "step": 11
    },
    {
      "epoch": 1.09,
      "learning_rate": 2e-05,
      "loss": 0.2841,
      "step": 12
    },
    {
      "epoch": 1.18,
      "learning_rate": 2e-05,
      "loss": 0.3026,
      "step": 13
    },
    {
      "epoch": 1.27,
      "learning_rate": 2e-05,
      "loss": 0.3741,
      "step": 14
    },
    {
      "epoch": 1.36,
      "learning_rate": 2e-05,
      "loss": 0.3297,
      "step": 15
    },
    {
      "epoch": 1.45,
      "learning_rate": 2e-05,
      "loss": 0.2938,
      "step": 16
    },
    {
      "epoch": 1.55,
      "learning_rate": 2e-05,
      "loss": 0.3768,
      "step": 17
    },
    {
      "epoch": 1.64,
      "learning_rate": 2e-05,
      "loss": 0.4006,
      "step": 18
    },
    {
      "epoch": 1.73,
      "learning_rate": 2e-05,
      "loss": 0.3016,
      "step": 19
    },
    {
      "epoch": 1.82,
      "learning_rate": 2e-05,
      "loss": 0.421,
      "step": 20
    },
    {
      "epoch": 1.91,
      "learning_rate": 2e-05,
      "loss": 0.5055,
      "step": 21
    },
    {
      "epoch": 2.0,
      "learning_rate": 2e-05,
      "loss": 0.3077,
      "step": 22
    },
    {
      "epoch": 2.0,
      "step": 22,
      "total_flos": 631723622400.0,
      "train_loss": 0.6941351687366312,
      "train_runtime": 806.6235,
      "train_samples_per_second": 3.283,
      "train_steps_per_second": 0.027
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 22,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 50000,
  "total_flos": 631723622400.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}