trainer_state.json

{
  "best_metric": 0.88,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0578/checkpoints/checkpoint-2997",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 2997,
  "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": 2.2054,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7464,
      "eval_loss": 1.087607979774475,
      "eval_runtime": 6.8439,
      "eval_samples_per_second": 547.931,
      "eval_steps_per_second": 8.621,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.788385391235352,
      "learning_rate": 3e-05,
      "loss": 0.7794,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8250666666666666,
      "eval_loss": 0.6746430397033691,
      "eval_runtime": 5.9404,
      "eval_samples_per_second": 631.265,
      "eval_steps_per_second": 9.932,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 3e-05,
      "loss": 0.4798,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8504,
      "eval_loss": 0.5304605960845947,
      "eval_runtime": 6.7152,
      "eval_samples_per_second": 558.432,
      "eval_steps_per_second": 8.786,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.709794044494629,
      "learning_rate": 3e-05,
      "loss": 0.3406,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8549333333333333,
      "eval_loss": 0.513992965221405,
      "eval_runtime": 7.0615,
      "eval_samples_per_second": 531.049,
      "eval_steps_per_second": 8.355,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.20353016257286072,
      "learning_rate": 3e-05,
      "loss": 0.2606,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8621333333333333,
      "eval_loss": 0.4979492127895355,
      "eval_runtime": 6.0024,
      "eval_samples_per_second": 624.747,
      "eval_steps_per_second": 9.829,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.4197559952735901,
      "learning_rate": 3e-05,
      "loss": 0.2008,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8656,
      "eval_loss": 0.48157310485839844,
      "eval_runtime": 6.0025,
      "eval_samples_per_second": 624.745,
      "eval_steps_per_second": 9.829,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.3616397380828857,
      "learning_rate": 3e-05,
      "loss": 0.1479,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8645333333333334,
      "eval_loss": 0.46858304738998413,
      "eval_runtime": 6.0427,
      "eval_samples_per_second": 620.588,
      "eval_steps_per_second": 9.764,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.72487211227417,
      "learning_rate": 3e-05,
      "loss": 0.1102,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8757333333333334,
      "eval_loss": 0.44478893280029297,
      "eval_runtime": 6.0144,
      "eval_samples_per_second": 623.501,
      "eval_steps_per_second": 9.81,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": Infinity,
      "learning_rate": 3e-05,
      "loss": 0.0891,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.88,
      "eval_loss": 0.4719252586364746,
      "eval_runtime": 6.2827,
      "eval_samples_per_second": 596.878,
      "eval_steps_per_second": 9.391,
      "step": 2997
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.482671885603328e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}