trainer_state.json

{
  "best_metric": 0.8725333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0080/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "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": 3e-05,
      "loss": 2.4074,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7157333333333333,
      "eval_loss": 1.2895668745040894,
      "eval_runtime": 7.8407,
      "eval_samples_per_second": 478.273,
      "eval_steps_per_second": 7.525,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 3e-05,
      "loss": 0.9091,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8138666666666666,
      "eval_loss": 0.7191084623336792,
      "eval_runtime": 4.5421,
      "eval_samples_per_second": 825.615,
      "eval_steps_per_second": 12.99,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.7370880842208862,
      "learning_rate": 3e-05,
      "loss": 0.5321,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8488,
      "eval_loss": 0.5471316576004028,
      "eval_runtime": 4.6801,
      "eval_samples_per_second": 801.262,
      "eval_steps_per_second": 12.607,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 9.050636291503906,
      "learning_rate": 3e-05,
      "loss": 0.3748,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8498666666666667,
      "eval_loss": 0.5332399606704712,
      "eval_runtime": 4.543,
      "eval_samples_per_second": 825.437,
      "eval_steps_per_second": 12.987,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 9.374919891357422,
      "learning_rate": 3e-05,
      "loss": 0.2718,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8632,
      "eval_loss": 0.4644162058830261,
      "eval_runtime": 4.4973,
      "eval_samples_per_second": 833.841,
      "eval_steps_per_second": 13.119,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.7333420515060425,
      "learning_rate": 3e-05,
      "loss": 0.2022,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8725333333333334,
      "eval_loss": 0.4472818374633789,
      "eval_runtime": 4.9207,
      "eval_samples_per_second": 762.082,
      "eval_steps_per_second": 11.99,
      "step": 1998
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}