trainer_state.json

{
  "best_metric": 0.9138666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0099/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": 4.850126564054637e-05,
      "loss": 1.8485,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8048,
      "eval_loss": 0.7771739363670349,
      "eval_runtime": 6.0921,
      "eval_samples_per_second": 615.55,
      "eval_steps_per_second": 9.685,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 4.4184758936602036e-05,
      "loss": 0.5398,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8616,
      "eval_loss": 0.49358242750167847,
      "eval_runtime": 5.538,
      "eval_samples_per_second": 677.137,
      "eval_steps_per_second": 10.654,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 20.573579788208008,
      "learning_rate": 3.7545362870758595e-05,
      "loss": 0.2906,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8789333333333333,
      "eval_loss": 0.4270245134830475,
      "eval_runtime": 5.3511,
      "eval_samples_per_second": 700.788,
      "eval_steps_per_second": 11.026,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 2.9418606892278543e-05,
      "loss": 0.1725,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.888,
      "eval_loss": 0.3886793553829193,
      "eval_runtime": 5.3664,
      "eval_samples_per_second": 698.797,
      "eval_steps_per_second": 10.994,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4165939688682556,
      "learning_rate": 2.073624094068776e-05,
      "loss": 0.0913,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9016,
      "eval_loss": 0.36575794219970703,
      "eval_runtime": 5.7956,
      "eval_samples_per_second": 647.043,
      "eval_steps_per_second": 10.18,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.5269932150840759,
      "learning_rate": 1.2568147258031899e-05,
      "loss": 0.0447,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9064,
      "eval_loss": 0.3543068468570709,
      "eval_runtime": 5.4684,
      "eval_samples_per_second": 685.753,
      "eval_steps_per_second": 10.789,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.07829136401414871,
      "learning_rate": 5.8995184906632666e-06,
      "loss": 0.0202,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9138666666666667,
      "eval_loss": 0.34335511922836304,
      "eval_runtime": 5.3779,
      "eval_samples_per_second": 697.296,
      "eval_steps_per_second": 10.971,
      "step": 2331
    }
  ],
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}