trainer_state.json

{
  "best_metric": 0.4392,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0471/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": 0.00028885116082809934,
      "loss": 3.3591,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.22026666666666667,
      "eval_loss": 2.957254648208618,
      "eval_runtime": 7.262,
      "eval_samples_per_second": 516.389,
      "eval_steps_per_second": 8.125,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 27.896129608154297,
      "learning_rate": 0.0002565651599376707,
      "loss": 2.8386,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2744,
      "eval_loss": 2.7134008407592773,
      "eval_runtime": 7.4183,
      "eval_samples_per_second": 505.504,
      "eval_steps_per_second": 7.953,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 26.669570922851562,
      "learning_rate": 0.00020805557946231208,
      "loss": 2.5855,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3024,
      "eval_loss": 2.535571336746216,
      "eval_runtime": 7.6165,
      "eval_samples_per_second": 492.352,
      "eval_steps_per_second": 7.746,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 25.023418426513672,
      "learning_rate": 0.00015070756328895485,
      "loss": 2.3572,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.34026666666666666,
      "eval_loss": 2.384960412979126,
      "eval_runtime": 7.4528,
      "eval_samples_per_second": 503.167,
      "eval_steps_per_second": 7.916,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 24.218364715576172,
      "learning_rate": 9.325182701893136e-05,
      "loss": 2.1065,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.38986666666666664,
      "eval_loss": 2.210226535797119,
      "eval_runtime": 7.2826,
      "eval_samples_per_second": 514.923,
      "eval_steps_per_second": 8.101,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 31.38882827758789,
      "learning_rate": 4.443548566180648e-05,
      "loss": 1.8481,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4112,
      "eval_loss": 2.1298584938049316,
      "eval_runtime": 7.3756,
      "eval_samples_per_second": 508.435,
      "eval_steps_per_second": 7.999,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 30.079545974731445,
      "learning_rate": 1.1690384667966174e-05,
      "loss": 1.6089,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4392,
      "eval_loss": 2.1000375747680664,
      "eval_runtime": 7.3462,
      "eval_samples_per_second": 510.47,
      "eval_steps_per_second": 8.031,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}