checkpoint-108/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.2057142857142855,
  "eval_steps": 500,
  "global_step": 108,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2,
      "learning_rate": 0.00012830188679245283,
      "loss": 1.7235,
      "step": 35
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.00019322033898305085,
      "loss": 1.226,
      "step": 70
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0001788135593220339,
      "loss": 0.72,
      "step": 105
    }
  ],
  "logging_steps": 35,
  "max_steps": 525,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 4746124116099072.0,
  "trial_name": null,
  "trial_params": null
}