trainer_state.json

{
  "best_metric": 1.438423752784729,
  "best_model_checkpoint": "tam_test_out_drug_data_new_run/checkpoint-1009",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1009,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5,
      "learning_rate": 3.026759167492567e-05,
      "loss": 1.7027,
      "step": 500
    },
    {
      "epoch": 0.99,
      "learning_rate": 5.351833498513379e-07,
      "loss": 1.4772,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4696363805449642,
      "eval_loss": 1.438423752784729,
      "eval_runtime": 24.4173,
      "eval_samples_per_second": 440.385,
      "eval_steps_per_second": 13.802,
      "step": 1009
    }
  ],
  "logging_steps": 500,
  "max_steps": 1009,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 3754801342955520.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}