trainer_state.json

{
  "best_metric": 0.9304,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0175/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": 0.6178297996520996,
      "learning_rate": 6.003003003003003e-05,
      "loss": 0.8049,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8696,
      "eval_loss": 0.44783639907836914,
      "eval_runtime": 7.7243,
      "eval_samples_per_second": 485.48,
      "eval_steps_per_second": 7.638,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 15.728116035461426,
      "learning_rate": 5.0030030030030024e-05,
      "loss": 0.2504,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8837333333333334,
      "eval_loss": 0.37587445974349976,
      "eval_runtime": 6.5281,
      "eval_samples_per_second": 574.443,
      "eval_steps_per_second": 9.038,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 4.006006006006006e-05,
      "loss": 0.1192,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8885333333333333,
      "eval_loss": 0.3987473249435425,
      "eval_runtime": 6.3724,
      "eval_samples_per_second": 588.474,
      "eval_steps_per_second": 9.259,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 3.0090090090090087e-05,
      "loss": 0.0726,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8981333333333333,
      "eval_loss": 0.38219866156578064,
      "eval_runtime": 6.3471,
      "eval_samples_per_second": 590.818,
      "eval_steps_per_second": 9.296,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0015386445447802544,
      "learning_rate": 2.009009009009009e-05,
      "loss": 0.0301,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9141333333333334,
      "eval_loss": 0.33128345012664795,
      "eval_runtime": 6.6528,
      "eval_samples_per_second": 563.668,
      "eval_steps_per_second": 8.868,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.00026579739642329514,
      "learning_rate": 1.0090090090090088e-05,
      "loss": 0.0084,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9256,
      "eval_loss": 0.2988567054271698,
      "eval_runtime": 6.3973,
      "eval_samples_per_second": 586.186,
      "eval_steps_per_second": 9.223,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.05677442252635956,
      "learning_rate": 9.009009009009008e-08,
      "loss": 0.0028,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9304,
      "eval_loss": 0.2706337869167328,
      "eval_runtime": 6.3445,
      "eval_samples_per_second": 591.066,
      "eval_steps_per_second": 9.299,
      "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
}