trainer_state.json

{
  "best_metric": 0.924,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0149/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": 9.623897705919314e-05,
      "loss": 0.8318,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.852,
      "eval_loss": 0.49417605996131897,
      "eval_runtime": 10.2249,
      "eval_samples_per_second": 366.751,
      "eval_steps_per_second": 5.77,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 8.548019870451392e-05,
      "loss": 0.2971,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8770666666666667,
      "eval_loss": 0.43968045711517334,
      "eval_runtime": 11.9423,
      "eval_samples_per_second": 314.009,
      "eval_steps_per_second": 4.94,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 10.209015846252441,
      "learning_rate": 6.929747794976645e-05,
      "loss": 0.1598,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8765333333333334,
      "eval_loss": 0.4451209604740143,
      "eval_runtime": 10.4425,
      "eval_samples_per_second": 359.111,
      "eval_steps_per_second": 5.65,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.06677880883216858,
      "learning_rate": 5.0176891109240266e-05,
      "loss": 0.0799,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8784,
      "eval_loss": 0.4533880352973938,
      "eval_runtime": 10.133,
      "eval_samples_per_second": 370.079,
      "eval_steps_per_second": 5.823,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.00044394255382940173,
      "learning_rate": 3.1029374200854165e-05,
      "loss": 0.0315,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9061333333333333,
      "eval_loss": 0.4109884202480316,
      "eval_runtime": 11.1542,
      "eval_samples_per_second": 336.197,
      "eval_steps_per_second": 5.289,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0011901799589395523,
      "learning_rate": 1.4769963101236895e-05,
      "loss": 0.0049,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9210666666666667,
      "eval_loss": 0.33768659830093384,
      "eval_runtime": 10.456,
      "eval_samples_per_second": 358.645,
      "eval_steps_per_second": 5.643,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0005972718936391175,
      "learning_rate": 3.87400575837657e-06,
      "loss": 0.0012,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.924,
      "eval_loss": 0.33039429783821106,
      "eval_runtime": 10.5493,
      "eval_samples_per_second": 355.473,
      "eval_steps_per_second": 5.593,
      "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
}