trainer_state.json

{
  "best_metric": 0.9218666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0009/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": 0.1698257029056549,
      "learning_rate": 0.0002851453301853628,
      "loss": 0.6069,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8608,
      "eval_loss": 0.4886538088321686,
      "eval_runtime": 11.8362,
      "eval_samples_per_second": 316.824,
      "eval_steps_per_second": 4.985,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00024368144166557083,
      "loss": 0.2707,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8805333333333333,
      "eval_loss": 0.41599205136299133,
      "eval_runtime": 11.8152,
      "eval_samples_per_second": 317.387,
      "eval_steps_per_second": 4.994,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 41.523250579833984,
      "learning_rate": 0.0001835752027932887,
      "loss": 0.1441,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8890666666666667,
      "eval_loss": 0.3925739526748657,
      "eval_runtime": 11.7438,
      "eval_samples_per_second": 319.317,
      "eval_steps_per_second": 5.024,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.03028416819870472,
      "learning_rate": 0.00011681898323495616,
      "loss": 0.058,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9026666666666666,
      "eval_loss": 0.38307926058769226,
      "eval_runtime": 11.582,
      "eval_samples_per_second": 323.779,
      "eval_steps_per_second": 5.094,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.001935211126692593,
      "learning_rate": 5.663467098540575e-05,
      "loss": 0.0122,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9189333333333334,
      "eval_loss": 0.3241618573665619,
      "eval_runtime": 11.7165,
      "eval_samples_per_second": 320.062,
      "eval_steps_per_second": 5.036,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.02023223042488098,
      "learning_rate": 1.4942507197812131e-05,
      "loss": 0.0025,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9216,
      "eval_loss": 0.31829577684402466,
      "eval_runtime": 11.7236,
      "eval_samples_per_second": 319.868,
      "eval_steps_per_second": 5.033,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.004329445771872997,
      "learning_rate": 1.3623114161753767e-10,
      "loss": 0.0014,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9218666666666666,
      "eval_loss": 0.31866517663002014,
      "eval_runtime": 11.9359,
      "eval_samples_per_second": 314.18,
      "eval_steps_per_second": 4.943,
      "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
}