trainer_state.json

{
  "best_metric": 0.8938666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0181/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": Infinity,
      "learning_rate": 4.8108254489432715e-05,
      "loss": 1.7849,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7946666666666666,
      "eval_loss": 0.7794867157936096,
      "eval_runtime": 6.877,
      "eval_samples_per_second": 545.298,
      "eval_steps_per_second": 8.579,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 12.895716667175293,
      "learning_rate": 4.2698504093359784e-05,
      "loss": 0.5024,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8488,
      "eval_loss": 0.5220008492469788,
      "eval_runtime": 6.4803,
      "eval_samples_per_second": 578.68,
      "eval_steps_per_second": 9.105,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.004144549369812,
      "learning_rate": 3.459431688641196e-05,
      "loss": 0.2538,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8706666666666667,
      "eval_loss": 0.43861880898475647,
      "eval_runtime": 6.5093,
      "eval_samples_per_second": 576.099,
      "eval_steps_per_second": 9.064,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 2.5058963771413844e-05,
      "loss": 0.1297,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8778666666666667,
      "eval_loss": 0.42082011699676514,
      "eval_runtime": 6.6259,
      "eval_samples_per_second": 565.962,
      "eval_steps_per_second": 8.904,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.21301676332950592,
      "learning_rate": 1.5487416222190068e-05,
      "loss": 0.0586,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8874666666666666,
      "eval_loss": 0.41124892234802246,
      "eval_runtime": 6.5464,
      "eval_samples_per_second": 572.833,
      "eval_steps_per_second": 9.013,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.09769764542579651,
      "learning_rate": 7.36407332135117e-06,
      "loss": 0.025,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8928,
      "eval_loss": 0.409416526556015,
      "eval_runtime": 6.7689,
      "eval_samples_per_second": 554.008,
      "eval_steps_per_second": 8.716,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.9263651371002197,
      "learning_rate": 1.925640387263111e-06,
      "loss": 0.0119,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8938666666666667,
      "eval_loss": 0.4095850884914398,
      "eval_runtime": 6.4461,
      "eval_samples_per_second": 581.743,
      "eval_steps_per_second": 9.153,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}