trainer_state.json

{
  "best_metric": 0.9008,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0339/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.6255983710289001,
      "learning_rate": 7e-05,
      "loss": 0.7601,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.872,
      "eval_loss": 0.4240235984325409,
      "eval_runtime": 7.0173,
      "eval_samples_per_second": 534.394,
      "eval_steps_per_second": 8.408,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 17.82039451599121,
      "learning_rate": 7e-05,
      "loss": 0.2735,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8688,
      "eval_loss": 0.43877318501472473,
      "eval_runtime": 6.7629,
      "eval_samples_per_second": 554.494,
      "eval_steps_per_second": 8.724,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.5660063028335571,
      "learning_rate": 7e-05,
      "loss": 0.1823,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8762666666666666,
      "eval_loss": 0.4474286437034607,
      "eval_runtime": 6.7331,
      "eval_samples_per_second": 556.946,
      "eval_steps_per_second": 8.763,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.264535665512085,
      "learning_rate": 7e-05,
      "loss": 0.1326,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8837333333333334,
      "eval_loss": 0.4557918310165405,
      "eval_runtime": 6.9178,
      "eval_samples_per_second": 542.079,
      "eval_steps_per_second": 8.529,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.03395872935652733,
      "learning_rate": 7e-05,
      "loss": 0.116,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8808,
      "eval_loss": 0.47939765453338623,
      "eval_runtime": 6.3113,
      "eval_samples_per_second": 594.17,
      "eval_steps_per_second": 9.348,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.5524324178695679,
      "learning_rate": 7e-05,
      "loss": 0.1046,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9008,
      "eval_loss": 0.41324207186698914,
      "eval_runtime": 6.3005,
      "eval_samples_per_second": 595.188,
      "eval_steps_per_second": 9.364,
      "step": 1998
    }
  ],
  "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": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}