trainer_state.json

{
  "best_metric": 0.9197333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0673/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": 1.1951229572296143,
      "learning_rate": 6.655434848731736e-05,
      "loss": 0.7479,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8605333333333334,
      "eval_loss": 0.47500964999198914,
      "eval_runtime": 10.6329,
      "eval_samples_per_second": 352.679,
      "eval_steps_per_second": 5.549,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.5193378925323486,
      "learning_rate": 5.685900305529986e-05,
      "loss": 0.2307,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8642666666666666,
      "eval_loss": 0.4655778706073761,
      "eval_runtime": 10.7039,
      "eval_samples_per_second": 350.34,
      "eval_steps_per_second": 5.512,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 4.288018105154487e-05,
      "loss": 0.1131,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8874666666666666,
      "eval_loss": 0.4125572144985199,
      "eval_runtime": 11.4634,
      "eval_samples_per_second": 327.129,
      "eval_steps_per_second": 5.147,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.05376364663243294,
      "learning_rate": 2.7303772261234784e-05,
      "loss": 0.0468,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8914666666666666,
      "eval_loss": 0.41394805908203125,
      "eval_runtime": 10.4292,
      "eval_samples_per_second": 359.566,
      "eval_steps_per_second": 5.657,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.05928726866841316,
      "learning_rate": 1.325169576262615e-05,
      "loss": 0.0117,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9090666666666667,
      "eval_loss": 0.3715037405490875,
      "eval_runtime": 10.7059,
      "eval_samples_per_second": 350.276,
      "eval_steps_per_second": 5.511,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.041666433215141296,
      "learning_rate": 3.5071376436769933e-06,
      "loss": 0.0025,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9165333333333333,
      "eval_loss": 0.334555059671402,
      "eval_runtime": 10.861,
      "eval_samples_per_second": 345.272,
      "eval_steps_per_second": 5.432,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.000587108894251287,
      "learning_rate": 1.2714900776311653e-10,
      "loss": 0.0009,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9197333333333333,
      "eval_loss": 0.3343414068222046,
      "eval_runtime": 10.4128,
      "eval_samples_per_second": 360.132,
      "eval_steps_per_second": 5.666,
      "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
}