trainer_state.json

{
  "best_metric": 0.8696,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0385/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": 78.024658203125,
      "learning_rate": 0.000225,
      "loss": 1.6237,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8181333333333334,
      "eval_loss": 0.6017518043518066,
      "eval_runtime": 7.3986,
      "eval_samples_per_second": 506.856,
      "eval_steps_per_second": 7.975,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 34.678443908691406,
      "learning_rate": 7.500000000000002e-05,
      "loss": 0.4479,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8685333333333334,
      "eval_loss": 0.41984492540359497,
      "eval_runtime": 7.4055,
      "eval_samples_per_second": 506.38,
      "eval_steps_per_second": 7.967,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 57.874698638916016,
      "learning_rate": 0.0,
      "loss": 0.28,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8696,
      "eval_loss": 0.4183524549007416,
      "eval_runtime": 7.4289,
      "eval_samples_per_second": 504.785,
      "eval_steps_per_second": 7.942,
      "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
}