trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5333333333333333,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.11,
      "learning_rate": 4.4983991462113126e-05,
      "loss": 2.4818,
      "step": 100
    },
    {
      "epoch": 0.21,
      "learning_rate": 3.970117395944504e-05,
      "loss": 2.2118,
      "step": 200
    },
    {
      "epoch": 0.32,
      "learning_rate": 3.436499466382071e-05,
      "loss": 2.1462,
      "step": 300
    },
    {
      "epoch": 0.43,
      "learning_rate": 2.9028815368196372e-05,
      "loss": 2.1053,
      "step": 400
    },
    {
      "epoch": 0.53,
      "learning_rate": 2.3692636072572037e-05,
      "loss": 2.075,
      "step": 500
    }
  ],
  "max_steps": 937,
  "num_train_epochs": 1,
  "total_flos": 3.07105194049536e+17,
  "trial_name": null,
  "trial_params": null
}