trainer_state.json

{
  "best_metric": 0.3248,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0903/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": Infinity,
      "learning_rate": 0.0005,
      "loss": 3.5141,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.17226666666666668,
      "eval_loss": 3.1767780780792236,
      "eval_runtime": 12.5909,
      "eval_samples_per_second": 297.834,
      "eval_steps_per_second": 4.686,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 38.90705108642578,
      "learning_rate": 0.0005,
      "loss": 3.0231,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2192,
      "eval_loss": 2.985227346420288,
      "eval_runtime": 12.4325,
      "eval_samples_per_second": 301.628,
      "eval_steps_per_second": 4.746,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 16.428234100341797,
      "learning_rate": 0.0005,
      "loss": 2.8625,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.23413333333333333,
      "eval_loss": 2.7886550426483154,
      "eval_runtime": 13.4669,
      "eval_samples_per_second": 278.46,
      "eval_steps_per_second": 4.381,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 30.64426612854004,
      "learning_rate": 0.0005,
      "loss": 2.7302,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.24506666666666665,
      "eval_loss": 2.7941906452178955,
      "eval_runtime": 12.3084,
      "eval_samples_per_second": 304.669,
      "eval_steps_per_second": 4.793,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 18.2420711517334,
      "learning_rate": 0.0005,
      "loss": 2.6253,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.2826666666666667,
      "eval_loss": 2.6274068355560303,
      "eval_runtime": 12.254,
      "eval_samples_per_second": 306.023,
      "eval_steps_per_second": 4.815,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 18.516616821289062,
      "learning_rate": 0.0005,
      "loss": 2.5411,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.308,
      "eval_loss": 2.5510852336883545,
      "eval_runtime": 12.3005,
      "eval_samples_per_second": 304.866,
      "eval_steps_per_second": 4.797,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 20.42148208618164,
      "learning_rate": 0.0005,
      "loss": 2.4564,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.3248,
      "eval_loss": 2.4950947761535645,
      "eval_runtime": 12.4738,
      "eval_samples_per_second": 300.63,
      "eval_steps_per_second": 4.73,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}