trainer_state.json

{
  "best_metric": 0.8744,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0710/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": 2.967390537261963,
      "learning_rate": 7e-05,
      "loss": 1.7045,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8037333333333333,
      "eval_loss": 0.727149248123169,
      "eval_runtime": 7.9396,
      "eval_samples_per_second": 472.313,
      "eval_steps_per_second": 7.431,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 7e-05,
      "loss": 0.5052,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8450666666666666,
      "eval_loss": 0.5238425135612488,
      "eval_runtime": 6.2323,
      "eval_samples_per_second": 601.701,
      "eval_steps_per_second": 9.467,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.901103496551514,
      "learning_rate": 7e-05,
      "loss": 0.3018,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8544,
      "eval_loss": 0.49437016248703003,
      "eval_runtime": 6.1831,
      "eval_samples_per_second": 606.493,
      "eval_steps_per_second": 9.542,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.3923960328102112,
      "learning_rate": 7e-05,
      "loss": 0.2035,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8586666666666667,
      "eval_loss": 0.48133137822151184,
      "eval_runtime": 6.6514,
      "eval_samples_per_second": 563.795,
      "eval_steps_per_second": 8.87,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.10363636165857315,
      "learning_rate": 7e-05,
      "loss": 0.1478,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8738666666666667,
      "eval_loss": 0.44488534331321716,
      "eval_runtime": 6.5195,
      "eval_samples_per_second": 575.196,
      "eval_steps_per_second": 9.05,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.2173660695552826,
      "learning_rate": 7e-05,
      "loss": 0.1132,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8677333333333334,
      "eval_loss": 0.459074467420578,
      "eval_runtime": 6.4658,
      "eval_samples_per_second": 579.97,
      "eval_steps_per_second": 9.125,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.08922484517097473,
      "learning_rate": 7e-05,
      "loss": 0.086,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8744,
      "eval_loss": 0.4707986116409302,
      "eval_runtime": 6.4871,
      "eval_samples_per_second": 578.07,
      "eval_steps_per_second": 9.095,
      "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
}