trainer_state.json

{
  "best_metric": 0.9213333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0485/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": 2.082667827606201,
      "learning_rate": 9.504844339512095e-05,
      "loss": 1.7597,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8186666666666667,
      "eval_loss": 0.6619030833244324,
      "eval_runtime": 7.6358,
      "eval_samples_per_second": 491.105,
      "eval_steps_per_second": 7.727,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 8.122714722185695e-05,
      "loss": 0.4848,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8741333333333333,
      "eval_loss": 0.43043020367622375,
      "eval_runtime": 7.4765,
      "eval_samples_per_second": 501.571,
      "eval_steps_per_second": 7.891,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 9.98570728302002,
      "learning_rate": 6.119173426442958e-05,
      "loss": 0.2658,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8912,
      "eval_loss": 0.38404929637908936,
      "eval_runtime": 9.9107,
      "eval_samples_per_second": 378.381,
      "eval_steps_per_second": 5.953,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.0082277059555054,
      "learning_rate": 3.893966107831872e-05,
      "loss": 0.1405,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9026666666666666,
      "eval_loss": 0.32970333099365234,
      "eval_runtime": 6.8722,
      "eval_samples_per_second": 545.675,
      "eval_steps_per_second": 8.585,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.058094918727874756,
      "learning_rate": 1.887822366180192e-05,
      "loss": 0.0706,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9165333333333333,
      "eval_loss": 0.30340856313705444,
      "eval_runtime": 6.8836,
      "eval_samples_per_second": 544.776,
      "eval_steps_per_second": 8.571,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.8742733597755432,
      "learning_rate": 4.980835732604045e-06,
      "loss": 0.0329,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9197333333333333,
      "eval_loss": 0.29066774249076843,
      "eval_runtime": 6.8218,
      "eval_samples_per_second": 549.712,
      "eval_steps_per_second": 8.649,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.4618031978607178,
      "learning_rate": 4.5410380539179233e-11,
      "loss": 0.0165,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9213333333333333,
      "eval_loss": 0.28699791431427,
      "eval_runtime": 6.9092,
      "eval_samples_per_second": 542.756,
      "eval_steps_per_second": 8.539,
      "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
}