trainer_state.json

{
  "best_metric": 0.9325333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0232/checkpoints/checkpoint-1665",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1665,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": NaN,
      "learning_rate": 9e-05,
      "loss": 0.6528,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9216,
      "eval_loss": 0.27261781692504883,
      "eval_runtime": 6.3226,
      "eval_samples_per_second": 593.108,
      "eval_steps_per_second": 9.332,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.11265988647937775,
      "learning_rate": 9e-05,
      "loss": 0.1581,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9216,
      "eval_loss": 0.2705393433570862,
      "eval_runtime": 7.3279,
      "eval_samples_per_second": 511.742,
      "eval_steps_per_second": 8.051,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 18.514606475830078,
      "learning_rate": 9e-05,
      "loss": 0.0926,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9184,
      "eval_loss": 0.2963384687900543,
      "eval_runtime": 5.8699,
      "eval_samples_per_second": 638.855,
      "eval_steps_per_second": 10.051,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.10436767339706421,
      "learning_rate": 9e-05,
      "loss": 0.0583,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9144,
      "eval_loss": 0.34549447894096375,
      "eval_runtime": 10.2009,
      "eval_samples_per_second": 367.616,
      "eval_steps_per_second": 5.784,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.1349575519561768,
      "learning_rate": 9e-05,
      "loss": 0.049,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9325333333333333,
      "eval_loss": 0.28770098090171814,
      "eval_runtime": 5.9309,
      "eval_samples_per_second": 632.277,
      "eval_steps_per_second": 9.948,
      "step": 1665
    }
  ],
  "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": 8.2370660311296e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}