trainer_state.json

{
  "best_metric": 0.41173333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0618/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": 0.0003,
      "loss": 3.3213,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.21626666666666666,
      "eval_loss": 2.9545648097991943,
      "eval_runtime": 14.6522,
      "eval_samples_per_second": 255.934,
      "eval_steps_per_second": 4.027,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 30.66193389892578,
      "learning_rate": 0.0003,
      "loss": 2.7892,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2754666666666667,
      "eval_loss": 2.652904748916626,
      "eval_runtime": 12.8217,
      "eval_samples_per_second": 292.472,
      "eval_steps_per_second": 4.602,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 32.24341583251953,
      "learning_rate": 0.0003,
      "loss": 2.5766,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.29786666666666667,
      "eval_loss": 2.5212490558624268,
      "eval_runtime": 12.9724,
      "eval_samples_per_second": 289.076,
      "eval_steps_per_second": 4.548,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 30.879772186279297,
      "learning_rate": 0.0003,
      "loss": 2.4032,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.33653333333333335,
      "eval_loss": 2.4007644653320312,
      "eval_runtime": 13.1718,
      "eval_samples_per_second": 284.699,
      "eval_steps_per_second": 4.479,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 30.149126052856445,
      "learning_rate": 0.0003,
      "loss": 2.2609,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3616,
      "eval_loss": 2.346041202545166,
      "eval_runtime": 12.1806,
      "eval_samples_per_second": 307.866,
      "eval_steps_per_second": 4.844,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 28.080188751220703,
      "learning_rate": 0.0003,
      "loss": 2.1315,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.39226666666666665,
      "eval_loss": 2.2028253078460693,
      "eval_runtime": 12.2067,
      "eval_samples_per_second": 307.209,
      "eval_steps_per_second": 4.833,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 23.871599197387695,
      "learning_rate": 0.0003,
      "loss": 1.9943,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.41173333333333334,
      "eval_loss": 2.127363920211792,
      "eval_runtime": 12.4052,
      "eval_samples_per_second": 302.293,
      "eval_steps_per_second": 4.756,
      "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
}