trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0001394353204167889,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 50.64784240722656,
      "learning_rate": 0.004999860564679583,
      "loss": 1.2003,
      "step": 100
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.446443557739258,
      "learning_rate": 0.004999721129359167,
      "loss": 1.8397,
      "step": 200
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.034487757831811905,
      "learning_rate": 0.0049995816940387496,
      "loss": 0.6524,
      "step": 300
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.0324031226336956,
      "learning_rate": 0.004999442258718333,
      "loss": 0.0663,
      "step": 400
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.10035407543182373,
      "learning_rate": 0.0049993028233979156,
      "loss": 0.0657,
      "step": 500
    }
  ],
  "logging_steps": 100,
  "max_steps": 3585892,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 735868440576000.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}