trainer_state.json

{
  "best_metric": 0.8650666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0625/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": 3e-05,
      "loss": 2.364,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.736,
      "eval_loss": 1.149181604385376,
      "eval_runtime": 6.6492,
      "eval_samples_per_second": 563.975,
      "eval_steps_per_second": 8.873,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.860910415649414,
      "learning_rate": 3e-05,
      "loss": 0.8006,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8304,
      "eval_loss": 0.6267068386077881,
      "eval_runtime": 5.8271,
      "eval_samples_per_second": 643.542,
      "eval_steps_per_second": 10.125,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 21.619518280029297,
      "learning_rate": 3e-05,
      "loss": 0.4568,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8517333333333333,
      "eval_loss": 0.5121111273765564,
      "eval_runtime": 5.8093,
      "eval_samples_per_second": 645.515,
      "eval_steps_per_second": 10.156,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 24.975421905517578,
      "learning_rate": 3e-05,
      "loss": 0.3005,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8568,
      "eval_loss": 0.4857499897480011,
      "eval_runtime": 5.8872,
      "eval_samples_per_second": 636.98,
      "eval_steps_per_second": 10.022,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.19912433624267578,
      "learning_rate": 3e-05,
      "loss": 0.2024,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8632,
      "eval_loss": 0.44575273990631104,
      "eval_runtime": 5.8188,
      "eval_samples_per_second": 644.457,
      "eval_steps_per_second": 10.139,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.986217021942139,
      "learning_rate": 3e-05,
      "loss": 0.138,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8608,
      "eval_loss": 0.47788184881210327,
      "eval_runtime": 5.8606,
      "eval_samples_per_second": 639.868,
      "eval_steps_per_second": 10.067,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.3395631313323975,
      "learning_rate": 3e-05,
      "loss": 0.1001,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8650666666666667,
      "eval_loss": 0.4900665879249573,
      "eval_runtime": 6.1911,
      "eval_samples_per_second": 605.707,
      "eval_steps_per_second": 9.53,
      "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
}