trainer_state.json

{
  "best_metric": 0.9312,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0060/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": 25.441545486450195,
      "learning_rate": 6.655434848731736e-05,
      "loss": 0.8226,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.852,
      "eval_loss": 0.49567532539367676,
      "eval_runtime": 12.0063,
      "eval_samples_per_second": 312.336,
      "eval_steps_per_second": 4.914,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 5.689582334051816e-05,
      "loss": 0.2921,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8730666666666667,
      "eval_loss": 0.43854373693466187,
      "eval_runtime": 12.4942,
      "eval_samples_per_second": 300.139,
      "eval_steps_per_second": 4.722,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 4.2926133804308225e-05,
      "loss": 0.145,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9066666666666666,
      "eval_loss": 0.33725813031196594,
      "eval_runtime": 11.4379,
      "eval_samples_per_second": 327.857,
      "eval_steps_per_second": 5.158,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.4224909543991089,
      "learning_rate": 2.7349795747191676e-05,
      "loss": 0.0645,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9056,
      "eval_loss": 0.335310161113739,
      "eval_runtime": 11.4449,
      "eval_samples_per_second": 327.657,
      "eval_steps_per_second": 5.155,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.012812451459467411,
      "learning_rate": 1.32886744659418e-05,
      "loss": 0.0192,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9285333333333333,
      "eval_loss": 0.27810919284820557,
      "eval_runtime": 11.4654,
      "eval_samples_per_second": 327.07,
      "eval_steps_per_second": 5.146,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.028930872678756714,
      "learning_rate": 3.5277474786457087e-06,
      "loss": 0.0045,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9282666666666667,
      "eval_loss": 0.28980451822280884,
      "eval_runtime": 11.3603,
      "eval_samples_per_second": 330.098,
      "eval_steps_per_second": 5.194,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.00750689534470439,
      "learning_rate": 2.8608505094840364e-10,
      "loss": 0.0015,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9312,
      "eval_loss": 0.2865731716156006,
      "eval_runtime": 11.4706,
      "eval_samples_per_second": 326.923,
      "eval_steps_per_second": 5.144,
      "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
}