trainer_state.json

{
  "best_metric": 0.4512,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0084/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 18.685924530029297,
      "learning_rate": 0.0003,
      "loss": 3.1978,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.24133333333333334,
      "eval_loss": 2.7559802532196045,
      "eval_runtime": 6.0575,
      "eval_samples_per_second": 619.068,
      "eval_steps_per_second": 9.74,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 24.263347625732422,
      "learning_rate": 0.0003,
      "loss": 2.5757,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3128,
      "eval_loss": 2.4831840991973877,
      "eval_runtime": 6.0512,
      "eval_samples_per_second": 619.707,
      "eval_steps_per_second": 9.75,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 20.477128982543945,
      "learning_rate": 0.0003,
      "loss": 2.3473,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.33253333333333335,
      "eval_loss": 2.427973508834839,
      "eval_runtime": 6.2135,
      "eval_samples_per_second": 603.525,
      "eval_steps_per_second": 9.495,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 69.66590118408203,
      "learning_rate": 0.0003,
      "loss": 2.1835,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.39226666666666665,
      "eval_loss": 2.192883014678955,
      "eval_runtime": 6.948,
      "eval_samples_per_second": 539.725,
      "eval_steps_per_second": 8.492,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 25.36086082458496,
      "learning_rate": 0.0003,
      "loss": 2.0383,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.39013333333333333,
      "eval_loss": 2.119925022125244,
      "eval_runtime": 5.8145,
      "eval_samples_per_second": 644.935,
      "eval_steps_per_second": 10.147,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 29.856386184692383,
      "learning_rate": 0.0003,
      "loss": 1.9427,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4384,
      "eval_loss": 1.9601616859436035,
      "eval_runtime": 5.8386,
      "eval_samples_per_second": 642.275,
      "eval_steps_per_second": 10.105,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0003,
      "loss": 1.8113,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4512,
      "eval_loss": 1.9212796688079834,
      "eval_runtime": 6.1126,
      "eval_samples_per_second": 613.483,
      "eval_steps_per_second": 9.652,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}