trainer_state.json

{
  "best_metric": 0.9018666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0253/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.11369079351425171,
      "learning_rate": 9.510675611750168e-05,
      "loss": 0.9823,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.792,
      "eval_loss": 0.677823543548584,
      "eval_runtime": 4.3713,
      "eval_samples_per_second": 857.876,
      "eval_steps_per_second": 13.497,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.081008434295654,
      "learning_rate": 8.127974762931168e-05,
      "loss": 0.3381,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8152,
      "eval_loss": 0.5889031291007996,
      "eval_runtime": 4.0819,
      "eval_samples_per_second": 918.68,
      "eval_steps_per_second": 14.454,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 8.295758247375488,
      "learning_rate": 6.125740150220697e-05,
      "loss": 0.1768,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.84,
      "eval_loss": 0.5558581948280334,
      "eval_runtime": 4.4655,
      "eval_samples_per_second": 839.773,
      "eval_steps_per_second": 13.212,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.812685251235962,
      "learning_rate": 3.900538894462112e-05,
      "loss": 0.0813,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.876,
      "eval_loss": 0.48300766944885254,
      "eval_runtime": 4.4101,
      "eval_samples_per_second": 850.323,
      "eval_steps_per_second": 13.378,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.003228748682886362,
      "learning_rate": 1.893099394660879e-05,
      "loss": 0.0276,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8962666666666667,
      "eval_loss": 0.43016842007637024,
      "eval_runtime": 4.0005,
      "eval_samples_per_second": 937.39,
      "eval_steps_per_second": 14.748,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0004665844899136573,
      "learning_rate": 5.0101966338242764e-06,
      "loss": 0.0031,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9018666666666667,
      "eval_loss": 0.40105846524238586,
      "eval_runtime": 5.0225,
      "eval_samples_per_second": 746.645,
      "eval_steps_per_second": 11.747,
      "step": 1998
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}