trainer_state.json

{
  "best_metric": 0.9266666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0289/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": 0.5491799116134644,
      "learning_rate": 2.574002574002574e-05,
      "loss": 0.7928,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8829333333333333,
      "eval_loss": 0.3706382215023041,
      "eval_runtime": 6.8438,
      "eval_samples_per_second": 547.942,
      "eval_steps_per_second": 8.621,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8332797288894653,
      "learning_rate": 2.1454311454311453e-05,
      "loss": 0.1972,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9066666666666666,
      "eval_loss": 0.30464258790016174,
      "eval_runtime": 7.2279,
      "eval_samples_per_second": 518.824,
      "eval_steps_per_second": 8.163,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 16.48934555053711,
      "learning_rate": 1.716859716859717e-05,
      "loss": 0.0809,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9085333333333333,
      "eval_loss": 0.2975218892097473,
      "eval_runtime": 6.7833,
      "eval_samples_per_second": 552.828,
      "eval_steps_per_second": 8.698,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.9582719802856445,
      "learning_rate": 1.2882882882882882e-05,
      "loss": 0.0351,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9146666666666666,
      "eval_loss": 0.31390348076820374,
      "eval_runtime": 6.3753,
      "eval_samples_per_second": 588.203,
      "eval_steps_per_second": 9.254,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.28336700797080994,
      "learning_rate": 8.597168597168598e-06,
      "loss": 0.0153,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9216,
      "eval_loss": 0.2930420935153961,
      "eval_runtime": 6.4321,
      "eval_samples_per_second": 583.013,
      "eval_steps_per_second": 9.173,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.048467494547367096,
      "learning_rate": 4.311454311454311e-06,
      "loss": 0.0053,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9221333333333334,
      "eval_loss": 0.2949390411376953,
      "eval_runtime": 6.4441,
      "eval_samples_per_second": 581.929,
      "eval_steps_per_second": 9.156,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.45515891909599304,
      "learning_rate": 2.5740025740025742e-08,
      "loss": 0.0017,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9266666666666666,
      "eval_loss": 0.27855825424194336,
      "eval_runtime": 6.3981,
      "eval_samples_per_second": 586.112,
      "eval_steps_per_second": 9.221,
      "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
}