trainer_state.json

{
  "best_metric": 0.9197333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0581/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": 8.562221064561357e-05,
      "loss": 0.8118,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8224,
      "eval_loss": 0.6022288799285889,
      "eval_runtime": 6.1213,
      "eval_samples_per_second": 612.617,
      "eval_steps_per_second": 9.639,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 42.76079559326172,
      "learning_rate": 7.319906209778102e-05,
      "loss": 0.2861,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8530666666666666,
      "eval_loss": 0.48572176694869995,
      "eval_runtime": 5.8481,
      "eval_samples_per_second": 641.234,
      "eval_steps_per_second": 10.089,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.7260416150093079,
      "learning_rate": 5.5190743462682014e-05,
      "loss": 0.1515,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8733333333333333,
      "eval_loss": 0.4268069863319397,
      "eval_runtime": 5.2284,
      "eval_samples_per_second": 717.233,
      "eval_steps_per_second": 11.284,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.007694638799875975,
      "learning_rate": 3.516402310353216e-05,
      "loss": 0.0685,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8874666666666666,
      "eval_loss": 0.4286271631717682,
      "eval_runtime": 5.6204,
      "eval_samples_per_second": 667.209,
      "eval_steps_per_second": 10.497,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 14.104719161987305,
      "learning_rate": 1.708543859906803e-05,
      "loss": 0.0217,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9074666666666666,
      "eval_loss": 0.3864201605319977,
      "eval_runtime": 5.4527,
      "eval_samples_per_second": 687.731,
      "eval_steps_per_second": 10.82,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0007069775019772351,
      "learning_rate": 4.53567532968734e-06,
      "loss": 0.002,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9186666666666666,
      "eval_loss": 0.3491797149181366,
      "eval_runtime": 5.7712,
      "eval_samples_per_second": 649.773,
      "eval_steps_per_second": 10.223,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.007614858914166689,
      "learning_rate": 3.6782363693366185e-10,
      "loss": 0.0002,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9197333333333333,
      "eval_loss": 0.3475312292575836,
      "eval_runtime": 5.6427,
      "eval_samples_per_second": 664.576,
      "eval_steps_per_second": 10.456,
      "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
}