trainer_state.json

{
  "best_metric": 0.7914666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0929/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": 0.00047538820348083834,
      "loss": 2.732,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.44533333333333336,
      "eval_loss": 1.953277587890625,
      "eval_runtime": 5.2065,
      "eval_samples_per_second": 720.252,
      "eval_steps_per_second": 11.332,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.903118133544922,
      "learning_rate": 0.00040613573610928476,
      "loss": 1.4586,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5816,
      "eval_loss": 1.4130582809448242,
      "eval_runtime": 5.2066,
      "eval_samples_per_second": 720.234,
      "eval_steps_per_second": 11.332,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.677294969558716,
      "learning_rate": 0.0003059586713221479,
      "loss": 0.8227,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7066666666666667,
      "eval_loss": 0.983507513999939,
      "eval_runtime": 5.2631,
      "eval_samples_per_second": 712.509,
      "eval_steps_per_second": 11.21,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.043544381856918335,
      "learning_rate": 0.00019469830539159362,
      "loss": 0.3884,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7490666666666667,
      "eval_loss": 0.8428646922111511,
      "eval_runtime": 6.5632,
      "eval_samples_per_second": 571.367,
      "eval_steps_per_second": 8.99,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.9034634828567505,
      "learning_rate": 9.43911183090096e-05,
      "loss": 0.1174,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7664,
      "eval_loss": 0.8548819422721863,
      "eval_runtime": 5.2675,
      "eval_samples_per_second": 711.909,
      "eval_steps_per_second": 11.201,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.97165584564209,
      "learning_rate": 2.490417866302022e-05,
      "loss": 0.0171,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7882666666666667,
      "eval_loss": 0.804702639579773,
      "eval_runtime": 5.3184,
      "eval_samples_per_second": 705.093,
      "eval_steps_per_second": 11.093,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.12384853512048721,
      "learning_rate": 2.2705190269589615e-10,
      "loss": 0.0055,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7914666666666667,
      "eval_loss": 0.7969250082969666,
      "eval_runtime": 6.7306,
      "eval_samples_per_second": 557.155,
      "eval_steps_per_second": 8.766,
      "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
}