trainer_state.json

{
  "best_metric": 0.5146666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0986/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": 0.0004810825448943271,
      "loss": 3.2696,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.23386666666666667,
      "eval_loss": 2.905519962310791,
      "eval_runtime": 6.7393,
      "eval_samples_per_second": 556.436,
      "eval_steps_per_second": 8.755,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0004271931404387096,
      "loss": 2.7115,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2888,
      "eval_loss": 2.605640411376953,
      "eval_runtime": 6.2958,
      "eval_samples_per_second": 595.633,
      "eval_steps_per_second": 9.371,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0003464873897488322,
      "loss": 2.3456,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.37546666666666667,
      "eval_loss": 2.262648105621338,
      "eval_runtime": 6.431,
      "eval_samples_per_second": 583.111,
      "eval_steps_per_second": 9.174,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 17.352325439453125,
      "learning_rate": 0.00025088445554620134,
      "loss": 2.0381,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.4162666666666667,
      "eval_loss": 2.100248098373413,
      "eval_runtime": 7.193,
      "eval_samples_per_second": 521.341,
      "eval_steps_per_second": 8.202,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 14.627814292907715,
      "learning_rate": 0.00015514687100427082,
      "loss": 1.7128,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4664,
      "eval_loss": 1.9236559867858887,
      "eval_runtime": 6.257,
      "eval_samples_per_second": 599.326,
      "eval_steps_per_second": 9.429,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 13.077484130859375,
      "learning_rate": 7.384981550618446e-05,
      "loss": 1.3476,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.49946666666666667,
      "eval_loss": 1.849387526512146,
      "eval_runtime": 7.1773,
      "eval_samples_per_second": 522.483,
      "eval_steps_per_second": 8.22,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 16.784534454345703,
      "learning_rate": 1.9370028791882848e-05,
      "loss": 0.9811,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5146666666666667,
      "eval_loss": 1.8142151832580566,
      "eval_runtime": 6.3972,
      "eval_samples_per_second": 586.198,
      "eval_steps_per_second": 9.223,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}