trainer_state.json

{
  "best_metric": 0.8557333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0151/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": 0.00025727155727155724,
      "loss": 2.3539,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6581333333333333,
      "eval_loss": 1.189218282699585,
      "eval_runtime": 7.3627,
      "eval_samples_per_second": 509.323,
      "eval_steps_per_second": 8.013,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7091535925865173,
      "learning_rate": 0.0002144144144144144,
      "loss": 0.8994,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7317333333333333,
      "eval_loss": 0.8909555077552795,
      "eval_runtime": 7.491,
      "eval_samples_per_second": 500.601,
      "eval_steps_per_second": 7.876,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.21921859681606293,
      "learning_rate": 0.00017155727155727157,
      "loss": 0.4899,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7813333333333333,
      "eval_loss": 0.7530983686447144,
      "eval_runtime": 7.2479,
      "eval_samples_per_second": 517.389,
      "eval_steps_per_second": 8.14,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 7.485422134399414,
      "learning_rate": 0.0001287001287001287,
      "loss": 0.2692,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8058666666666666,
      "eval_loss": 0.6456692814826965,
      "eval_runtime": 7.505,
      "eval_samples_per_second": 499.664,
      "eval_steps_per_second": 7.861,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.6813873052597046,
      "learning_rate": 8.584298584298584e-05,
      "loss": 0.1257,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8192,
      "eval_loss": 0.6718268394470215,
      "eval_runtime": 9.6771,
      "eval_samples_per_second": 387.511,
      "eval_steps_per_second": 6.097,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0922919362783432,
      "learning_rate": 4.298584298584298e-05,
      "loss": 0.0513,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8450666666666666,
      "eval_loss": 0.587394118309021,
      "eval_runtime": 7.5058,
      "eval_samples_per_second": 499.612,
      "eval_steps_per_second": 7.861,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.07259310781955719,
      "learning_rate": 1.2870012870012868e-07,
      "loss": 0.018,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8557333333333333,
      "eval_loss": 0.5513054132461548,
      "eval_runtime": 7.4219,
      "eval_samples_per_second": 505.261,
      "eval_steps_per_second": 7.949,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}