trainer_state.json

{
  "best_metric": 0.8789333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0672/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": 5e-05,
      "loss": 1.8144,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8008,
      "eval_loss": 0.7732277512550354,
      "eval_runtime": 4.06,
      "eval_samples_per_second": 923.646,
      "eval_steps_per_second": 14.532,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 5e-05,
      "loss": 0.5471,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8514666666666667,
      "eval_loss": 0.528920590877533,
      "eval_runtime": 4.142,
      "eval_samples_per_second": 905.358,
      "eval_steps_per_second": 14.244,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5707992911338806,
      "learning_rate": 5e-05,
      "loss": 0.3195,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8650666666666667,
      "eval_loss": 0.472063273191452,
      "eval_runtime": 4.0154,
      "eval_samples_per_second": 933.907,
      "eval_steps_per_second": 14.693,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.17235355079174042,
      "learning_rate": 5e-05,
      "loss": 0.2176,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8701333333333333,
      "eval_loss": 0.44999960064888,
      "eval_runtime": 4.0502,
      "eval_samples_per_second": 925.876,
      "eval_steps_per_second": 14.567,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.20940488576889038,
      "learning_rate": 5e-05,
      "loss": 0.1586,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8650666666666667,
      "eval_loss": 0.4718398153781891,
      "eval_runtime": 4.0171,
      "eval_samples_per_second": 933.52,
      "eval_steps_per_second": 14.687,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 7.499838352203369,
      "learning_rate": 5e-05,
      "loss": 0.1132,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8789333333333333,
      "eval_loss": 0.4390992224216461,
      "eval_runtime": 5.2468,
      "eval_samples_per_second": 714.719,
      "eval_steps_per_second": 11.245,
      "step": 1998
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}