trainer_state.json

{
  "best_metric": 0.9141333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0675/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": 5.602887153625488,
      "learning_rate": 0.00042857142857142855,
      "loss": 0.7454,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8189333333333333,
      "eval_loss": 0.6322317123413086,
      "eval_runtime": 10.5085,
      "eval_samples_per_second": 356.856,
      "eval_steps_per_second": 5.615,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00035735735735735734,
      "loss": 0.3856,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8648,
      "eval_loss": 0.4554492235183716,
      "eval_runtime": 10.9575,
      "eval_samples_per_second": 342.231,
      "eval_steps_per_second": 5.384,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.01886007934808731,
      "learning_rate": 0.00028592878592878594,
      "loss": 0.2267,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.88,
      "eval_loss": 0.43331822752952576,
      "eval_runtime": 10.1773,
      "eval_samples_per_second": 368.466,
      "eval_steps_per_second": 5.797,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5305486917495728,
      "learning_rate": 0.0002145002145002145,
      "loss": 0.1242,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8856,
      "eval_loss": 0.4290192723274231,
      "eval_runtime": 10.2962,
      "eval_samples_per_second": 364.211,
      "eval_steps_per_second": 5.73,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.025812502950429916,
      "learning_rate": 0.00014307164307164307,
      "loss": 0.0589,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8952,
      "eval_loss": 0.42499813437461853,
      "eval_runtime": 10.262,
      "eval_samples_per_second": 365.425,
      "eval_steps_per_second": 5.749,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.12145310640335083,
      "learning_rate": 7.164307164307165e-05,
      "loss": 0.0174,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.912,
      "eval_loss": 0.3462689518928528,
      "eval_runtime": 10.5854,
      "eval_samples_per_second": 354.26,
      "eval_steps_per_second": 5.574,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0015372628113254905,
      "learning_rate": 2.1450021450021452e-07,
      "loss": 0.0037,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9141333333333334,
      "eval_loss": 0.3422955870628357,
      "eval_runtime": 10.497,
      "eval_samples_per_second": 357.245,
      "eval_steps_per_second": 5.621,
      "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
}