trainer_state.json

{
  "best_metric": 0.8498666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0042/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": 25.521907806396484,
      "learning_rate": 0.0005,
      "loss": 1.444,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8058666666666666,
      "eval_loss": 0.6533881425857544,
      "eval_runtime": 15.5624,
      "eval_samples_per_second": 240.966,
      "eval_steps_per_second": 3.791,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 64.43408203125,
      "learning_rate": 0.0005,
      "loss": 0.545,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8418666666666667,
      "eval_loss": 0.530379593372345,
      "eval_runtime": 15.6493,
      "eval_samples_per_second": 239.627,
      "eval_steps_per_second": 3.77,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 22.1973819732666,
      "learning_rate": 0.0005,
      "loss": 0.361,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8498666666666667,
      "eval_loss": 0.5243920087814331,
      "eval_runtime": 15.4195,
      "eval_samples_per_second": 243.199,
      "eval_steps_per_second": 3.826,
      "step": 999
    }
  ],
  "logging_steps": 500,
  "max_steps": 1332,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.4529287718656e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}