trainer_state.json

{
  "best_metric": 0.9328,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0075/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": 0.4210307002067566,
      "learning_rate": 7e-05,
      "loss": 0.7285,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9256,
      "eval_loss": 0.2696166932582855,
      "eval_runtime": 11.9749,
      "eval_samples_per_second": 313.155,
      "eval_steps_per_second": 4.927,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.838354229927063,
      "learning_rate": 7e-05,
      "loss": 0.1421,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9261333333333334,
      "eval_loss": 0.2714262902736664,
      "eval_runtime": 10.9988,
      "eval_samples_per_second": 340.946,
      "eval_steps_per_second": 5.364,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.3708128929138184,
      "learning_rate": 7e-05,
      "loss": 0.0707,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9277333333333333,
      "eval_loss": 0.27729591727256775,
      "eval_runtime": 10.9127,
      "eval_samples_per_second": 343.635,
      "eval_steps_per_second": 5.407,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": NaN,
      "learning_rate": 7e-05,
      "loss": 0.0526,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9301333333333334,
      "eval_loss": 0.26862162351608276,
      "eval_runtime": 10.8385,
      "eval_samples_per_second": 345.987,
      "eval_steps_per_second": 5.444,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": Infinity,
      "learning_rate": 7e-05,
      "loss": 0.0467,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9328,
      "eval_loss": 0.27663490176200867,
      "eval_runtime": 11.0499,
      "eval_samples_per_second": 339.37,
      "eval_steps_per_second": 5.339,
      "step": 1665
    }
  ],
  "logging_steps": 500,
  "max_steps": 1665,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 8.2370660311296e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}