checkpoint-3/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8,
  "eval_steps": 1,
  "global_step": 3,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 0.1357570467836887,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 1.7723,
      "step": 1
    },
    {
      "epoch": 0.26666666666666666,
      "eval_loss": 2.088420867919922,
      "eval_runtime": 6.4988,
      "eval_samples_per_second": 2.308,
      "eval_steps_per_second": 0.154,
      "step": 1
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 0.12515875774524352,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 1.8176,
      "step": 2
    },
    {
      "epoch": 0.5333333333333333,
      "eval_loss": 2.0871875286102295,
      "eval_runtime": 6.4536,
      "eval_samples_per_second": 2.324,
      "eval_steps_per_second": 0.155,
      "step": 2
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.126303798585066,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 1.8499,
      "step": 3
    },
    {
      "epoch": 0.8,
      "eval_loss": 2.087430000305176,
      "eval_runtime": 6.4444,
      "eval_samples_per_second": 2.328,
      "eval_steps_per_second": 0.155,
      "step": 3
    }
  ],
  "logging_steps": 1,
  "max_steps": 12,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 3,
  "total_flos": 4434721505280.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}