trainer_state.json

{
  "best_metric": 0.9490666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0890/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.293199360370636,
      "learning_rate": 8.888888888888889e-05,
      "loss": 0.5966,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9232,
      "eval_loss": 0.25856003165245056,
      "eval_runtime": 12.0251,
      "eval_samples_per_second": 311.849,
      "eval_steps_per_second": 4.906,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.11764486879110336,
      "learning_rate": 7.777777777777778e-05,
      "loss": 0.146,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9352,
      "eval_loss": 0.22148893773555756,
      "eval_runtime": 11.893,
      "eval_samples_per_second": 315.312,
      "eval_steps_per_second": 4.961,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.06396552175283432,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0796,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.932,
      "eval_loss": 0.23684926331043243,
      "eval_runtime": 11.9547,
      "eval_samples_per_second": 313.685,
      "eval_steps_per_second": 4.935,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.020424194633960724,
      "learning_rate": 5.555555555555556e-05,
      "loss": 0.0473,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9424,
      "eval_loss": 0.2288062870502472,
      "eval_runtime": 11.8217,
      "eval_samples_per_second": 317.213,
      "eval_steps_per_second": 4.991,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.020549962297081947,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.0267,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9410666666666667,
      "eval_loss": 0.23294219374656677,
      "eval_runtime": 11.9648,
      "eval_samples_per_second": 313.419,
      "eval_steps_per_second": 4.931,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.020066628232598305,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0118,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9381333333333334,
      "eval_loss": 0.23991215229034424,
      "eval_runtime": 11.9085,
      "eval_samples_per_second": 314.9,
      "eval_steps_per_second": 4.954,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 2.225558892225559e-05,
      "loss": 0.007,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9442666666666667,
      "eval_loss": 0.2286410629749298,
      "eval_runtime": 11.923,
      "eval_samples_per_second": 314.518,
      "eval_steps_per_second": 4.948,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.008122207596898079,
      "learning_rate": 1.1144477811144478e-05,
      "loss": 0.0025,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9490666666666666,
      "eval_loss": 0.2157890796661377,
      "eval_runtime": 11.7762,
      "eval_samples_per_second": 318.44,
      "eval_steps_per_second": 5.01,
      "step": 2664
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}