trainer_state.json

{
  "best_metric": 0.9005333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0549/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": 9.62613808353013e-05,
      "loss": 0.9807,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8098666666666666,
      "eval_loss": 0.6213652491569519,
      "eval_runtime": 13.8478,
      "eval_samples_per_second": 270.8,
      "eval_steps_per_second": 4.261,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.18266263604164124,
      "learning_rate": 8.548019870451392e-05,
      "loss": 0.3396,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8397333333333333,
      "eval_loss": 0.5285243988037109,
      "eval_runtime": 10.6936,
      "eval_samples_per_second": 350.677,
      "eval_steps_per_second": 5.517,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.09701050072908401,
      "learning_rate": 6.929747794976645e-05,
      "loss": 0.168,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8613333333333333,
      "eval_loss": 0.5127843022346497,
      "eval_runtime": 11.0903,
      "eval_samples_per_second": 338.132,
      "eval_steps_per_second": 5.32,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.002564536640420556,
      "learning_rate": 5.0176891109240266e-05,
      "loss": 0.0846,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8610666666666666,
      "eval_loss": 0.5694946646690369,
      "eval_runtime": 10.6925,
      "eval_samples_per_second": 350.713,
      "eval_steps_per_second": 5.518,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.013447609730064869,
      "learning_rate": 3.1029374200854165e-05,
      "loss": 0.0372,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8696,
      "eval_loss": 0.5628643035888672,
      "eval_runtime": 13.2422,
      "eval_samples_per_second": 283.185,
      "eval_steps_per_second": 4.455,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.10165601968765259,
      "learning_rate": 1.4769963101236895e-05,
      "loss": 0.0095,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8946666666666667,
      "eval_loss": 0.48835986852645874,
      "eval_runtime": 10.388,
      "eval_samples_per_second": 360.994,
      "eval_steps_per_second": 5.68,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0020811024587601423,
      "learning_rate": 3.87400575837657e-06,
      "loss": 0.0011,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9005333333333333,
      "eval_loss": 0.46969062089920044,
      "eval_runtime": 10.3833,
      "eval_samples_per_second": 361.156,
      "eval_steps_per_second": 5.682,
      "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
}