trainer_state.json

{
  "best_metric": 0.9248,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0103/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": 1.310793399810791,
      "learning_rate": 8.559608050575152e-05,
      "loss": 0.8624,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8128,
      "eval_loss": 0.639226496219635,
      "eval_runtime": 3.9765,
      "eval_samples_per_second": 943.032,
      "eval_steps_per_second": 14.837,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7200691103935242,
      "learning_rate": 7.31517728663805e-05,
      "loss": 0.3054,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8626666666666667,
      "eval_loss": 0.48905709385871887,
      "eval_runtime": 3.9921,
      "eval_samples_per_second": 939.354,
      "eval_steps_per_second": 14.779,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.10306277126073837,
      "learning_rate": 5.5131661351986275e-05,
      "loss": 0.1483,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8810666666666667,
      "eval_loss": 0.42686647176742554,
      "eval_runtime": 4.4782,
      "eval_samples_per_second": 837.383,
      "eval_steps_per_second": 13.175,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.012283108197152615,
      "learning_rate": 3.5104850050159014e-05,
      "loss": 0.0618,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8832,
      "eval_loss": 0.43109944462776184,
      "eval_runtime": 4.0425,
      "eval_samples_per_second": 927.652,
      "eval_steps_per_second": 14.595,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.263368129730225,
      "learning_rate": 1.703789455194791e-05,
      "loss": 0.0235,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9149333333333334,
      "eval_loss": 0.3406287729740143,
      "eval_runtime": 4.285,
      "eval_samples_per_second": 875.15,
      "eval_steps_per_second": 13.769,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.13513527810573578,
      "learning_rate": 4.509176970441849e-06,
      "loss": 0.0058,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9245333333333333,
      "eval_loss": 0.31397223472595215,
      "eval_runtime": 4.0189,
      "eval_samples_per_second": 933.086,
      "eval_steps_per_second": 14.681,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.005714153405278921,
      "learning_rate": 1.6347729569543557e-10,
      "loss": 0.0017,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9248,
      "eval_loss": 0.3171858787536621,
      "eval_runtime": 4.1475,
      "eval_samples_per_second": 904.163,
      "eval_steps_per_second": 14.225,
      "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
}