trainer_state.json

{
  "best_metric": 0.872,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0782/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": 5e-05,
      "loss": 0.8534,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8594666666666667,
      "eval_loss": 0.4509929120540619,
      "eval_runtime": 13.8354,
      "eval_samples_per_second": 271.043,
      "eval_steps_per_second": 4.264,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.2619872093200684,
      "learning_rate": 5e-05,
      "loss": 0.2753,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8613333333333333,
      "eval_loss": 0.46406546235084534,
      "eval_runtime": 13.1458,
      "eval_samples_per_second": 285.261,
      "eval_steps_per_second": 4.488,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 85.50745391845703,
      "learning_rate": 5e-05,
      "loss": 0.1542,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8704,
      "eval_loss": 0.465345174074173,
      "eval_runtime": 13.9624,
      "eval_samples_per_second": 268.578,
      "eval_steps_per_second": 4.226,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 19.20318031311035,
      "learning_rate": 5e-05,
      "loss": 0.1123,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8669333333333333,
      "eval_loss": 0.5019576549530029,
      "eval_runtime": 13.1322,
      "eval_samples_per_second": 285.557,
      "eval_steps_per_second": 4.493,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0023162481375038624,
      "learning_rate": 5e-05,
      "loss": 0.0813,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8477333333333333,
      "eval_loss": 0.5918921232223511,
      "eval_runtime": 12.707,
      "eval_samples_per_second": 295.112,
      "eval_steps_per_second": 4.643,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.015289923176169395,
      "learning_rate": 5e-05,
      "loss": 0.0694,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8682666666666666,
      "eval_loss": 0.5364126563072205,
      "eval_runtime": 13.0196,
      "eval_samples_per_second": 288.027,
      "eval_steps_per_second": 4.532,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.11781767755746841,
      "learning_rate": 5e-05,
      "loss": 0.0581,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.872,
      "eval_loss": 0.5657901167869568,
      "eval_runtime": 12.779,
      "eval_samples_per_second": 293.45,
      "eval_steps_per_second": 4.617,
      "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
}