run2/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.25,
      "grad_norm": 1.6990946531295776,
      "learning_rate": 0.00095,
      "loss": 0.6702,
      "step": 25
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.1647204160690308,
      "learning_rate": 0.000902,
      "loss": 0.3786,
      "step": 50
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.5457615852355957,
      "learning_rate": 0.000852,
      "loss": 0.3549,
      "step": 75
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.4196206331253052,
      "learning_rate": 0.0008020000000000001,
      "loss": 0.3115,
      "step": 100
    }
  ],
  "logging_steps": 25,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 100,
  "total_flos": 2.34945183744e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}