trainer_state.json

{
  "best_metric": 0.9352,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0247/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": 8.901763916015625,
      "learning_rate": 0.0002851453301853628,
      "loss": 0.6416,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8696,
      "eval_loss": 0.4261356592178345,
      "eval_runtime": 4.4359,
      "eval_samples_per_second": 845.381,
      "eval_steps_per_second": 13.301,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.08236990869045258,
      "learning_rate": 0.00024352347027881003,
      "loss": 0.2878,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8944,
      "eval_loss": 0.36080196499824524,
      "eval_runtime": 4.4528,
      "eval_samples_per_second": 842.173,
      "eval_steps_per_second": 13.25,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.3495333194732666,
      "learning_rate": 0.00018337814009344714,
      "loss": 0.1525,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8866666666666667,
      "eval_loss": 0.4295538663864136,
      "eval_runtime": 4.9359,
      "eval_samples_per_second": 759.737,
      "eval_steps_per_second": 11.953,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 10.540578842163086,
      "learning_rate": 0.00011662185990655284,
      "loss": 0.0763,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9005333333333333,
      "eval_loss": 0.3626863956451416,
      "eval_runtime": 4.3097,
      "eval_samples_per_second": 870.137,
      "eval_steps_per_second": 13.69,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.007647243794053793,
      "learning_rate": 5.6476529721189974e-05,
      "loss": 0.0249,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9258666666666666,
      "eval_loss": 0.32168835401535034,
      "eval_runtime": 4.3735,
      "eval_samples_per_second": 857.442,
      "eval_steps_per_second": 13.49,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.006878294050693512,
      "learning_rate": 1.4854669814637143e-05,
      "loss": 0.0063,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9325333333333333,
      "eval_loss": 0.2857852876186371,
      "eval_runtime": 4.3828,
      "eval_samples_per_second": 855.613,
      "eval_steps_per_second": 13.462,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.3623114161753767e-10,
      "loss": 0.0022,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9352,
      "eval_loss": 0.2867874205112457,
      "eval_runtime": 4.3883,
      "eval_samples_per_second": 854.547,
      "eval_steps_per_second": 13.445,
      "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
}