trainer_state.json

{
  "best_metric": 0.8632,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0926/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": 2.2914,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6424,
      "eval_loss": 1.2402721643447876,
      "eval_runtime": 8.0206,
      "eval_samples_per_second": 467.544,
      "eval_steps_per_second": 7.356,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.9408851861953735,
      "learning_rate": 0.0002144144144144144,
      "loss": 0.8692,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7565333333333333,
      "eval_loss": 0.7991046905517578,
      "eval_runtime": 6.7933,
      "eval_samples_per_second": 552.015,
      "eval_steps_per_second": 8.685,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00017168597168597168,
      "loss": 0.5126,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7834666666666666,
      "eval_loss": 0.712195873260498,
      "eval_runtime": 6.8492,
      "eval_samples_per_second": 547.51,
      "eval_steps_per_second": 8.614,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5007272362709045,
      "learning_rate": 0.0001288288288288288,
      "loss": 0.2935,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8197333333333333,
      "eval_loss": 0.6039271354675293,
      "eval_runtime": 6.852,
      "eval_samples_per_second": 547.285,
      "eval_steps_per_second": 8.611,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 10.887737274169922,
      "learning_rate": 8.597168597168596e-05,
      "loss": 0.1517,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8301333333333333,
      "eval_loss": 0.5912231802940369,
      "eval_runtime": 6.8937,
      "eval_samples_per_second": 543.974,
      "eval_steps_per_second": 8.559,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.893837928771973,
      "learning_rate": 4.311454311454311e-05,
      "loss": 0.0659,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8554666666666667,
      "eval_loss": 0.5332266092300415,
      "eval_runtime": 7.2847,
      "eval_samples_per_second": 514.781,
      "eval_steps_per_second": 8.099,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.13510097563266754,
      "learning_rate": 2.5740025740025737e-07,
      "loss": 0.0223,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8632,
      "eval_loss": 0.5030817985534668,
      "eval_runtime": 6.842,
      "eval_samples_per_second": 548.086,
      "eval_steps_per_second": 8.623,
      "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
}