trainer_state.json

{
  "best_metric": 0.8610666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0488/checkpoints/checkpoint-999",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 999,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 288.9173889160156,
      "learning_rate": 5e-05,
      "loss": 3.5909,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5728,
      "eval_loss": 2.778334379196167,
      "eval_runtime": 8.8706,
      "eval_samples_per_second": 422.747,
      "eval_steps_per_second": 6.651,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 154.06622314453125,
      "learning_rate": 5e-05,
      "loss": 1.475,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.804,
      "eval_loss": 0.7218507528305054,
      "eval_runtime": 8.6174,
      "eval_samples_per_second": 435.166,
      "eval_steps_per_second": 6.847,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 291.40185546875,
      "learning_rate": 5e-05,
      "loss": 0.5852,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8610666666666666,
      "eval_loss": 0.4731553792953491,
      "eval_runtime": 8.6818,
      "eval_samples_per_second": 431.939,
      "eval_steps_per_second": 6.796,
      "step": 999
    }
  ],
  "logging_steps": 500,
  "max_steps": 999,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.4529287718656e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}