trainer_state.json

{
  "best_metric": 0.5312,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0570/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": Infinity,
      "learning_rate": 0.0003,
      "loss": 3.2433,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.22506666666666666,
      "eval_loss": 2.90881085395813,
      "eval_runtime": 6.6978,
      "eval_samples_per_second": 559.881,
      "eval_steps_per_second": 8.809,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0003,
      "loss": 2.6182,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3408,
      "eval_loss": 2.396495819091797,
      "eval_runtime": 6.7288,
      "eval_samples_per_second": 557.309,
      "eval_steps_per_second": 8.768,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 16.869314193725586,
      "learning_rate": 0.0003,
      "loss": 2.1626,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4165333333333333,
      "eval_loss": 2.0774524211883545,
      "eval_runtime": 6.7002,
      "eval_samples_per_second": 559.685,
      "eval_steps_per_second": 8.806,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 16.119834899902344,
      "learning_rate": 0.0003,
      "loss": 1.7577,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.4669333333333333,
      "eval_loss": 1.879685878753662,
      "eval_runtime": 6.7205,
      "eval_samples_per_second": 557.996,
      "eval_steps_per_second": 8.779,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0003,
      "loss": 1.4116,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5312,
      "eval_loss": 1.651200771331787,
      "eval_runtime": 6.5915,
      "eval_samples_per_second": 568.917,
      "eval_steps_per_second": 8.951,
      "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
}