trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 6500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.15,
      "grad_norm": 0.9831129908561707,
      "learning_rate": 4.714201332828372e-05,
      "loss": 0.6011,
      "step": 1000
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.8143503665924072,
      "learning_rate": 3.921156116324626e-05,
      "loss": 0.58,
      "step": 2000
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9323546290397644,
      "learning_rate": 2.8025411603587355e-05,
      "loss": 0.5678,
      "step": 3000
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.5892429947853088,
      "learning_rate": 1.6146176705708874e-05,
      "loss": 0.5665,
      "step": 4000
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8226454257965088,
      "learning_rate": 6.311288588223915e-06,
      "loss": 0.564,
      "step": 5000
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.0760297775268555,
      "learning_rate": 7.351550661224088e-07,
      "loss": 0.5565,
      "step": 6000
    },
    {
      "epoch": 1.0,
      "step": 6500,
      "total_flos": 5.347603338846536e+17,
      "train_loss": 0.5714873938927284,
      "train_runtime": 18928.0922,
      "train_samples_per_second": 1.374,
      "train_steps_per_second": 0.343
    }
  ],
  "logging_steps": 1000,
  "max_steps": 6500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "total_flos": 5.347603338846536e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}