trainer_state.json

{
  "best_metric": 0.49146666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0398/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": 0.0005,
      "loss": 3.1832,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.22426666666666667,
      "eval_loss": 2.8455588817596436,
      "eval_runtime": 6.4304,
      "eval_samples_per_second": 583.163,
      "eval_steps_per_second": 9.175,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.6315,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3101333333333333,
      "eval_loss": 2.461355209350586,
      "eval_runtime": 6.259,
      "eval_samples_per_second": 599.133,
      "eval_steps_per_second": 9.426,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 16.108219146728516,
      "learning_rate": 0.0005,
      "loss": 2.2474,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3834666666666667,
      "eval_loss": 2.1859986782073975,
      "eval_runtime": 6.2775,
      "eval_samples_per_second": 597.368,
      "eval_steps_per_second": 9.399,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 13.056293487548828,
      "learning_rate": 0.0005,
      "loss": 1.9666,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.4125333333333333,
      "eval_loss": 2.0839362144470215,
      "eval_runtime": 7.2986,
      "eval_samples_per_second": 513.798,
      "eval_steps_per_second": 8.084,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 1.7627,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4416,
      "eval_loss": 1.9521427154541016,
      "eval_runtime": 6.3575,
      "eval_samples_per_second": 589.85,
      "eval_steps_per_second": 9.28,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 15.859006881713867,
      "learning_rate": 0.0005,
      "loss": 1.5481,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.46586666666666665,
      "eval_loss": 1.8925620317459106,
      "eval_runtime": 6.7214,
      "eval_samples_per_second": 557.916,
      "eval_steps_per_second": 8.778,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 18.078311920166016,
      "learning_rate": 0.0005,
      "loss": 1.3583,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.49146666666666666,
      "eval_loss": 1.8393820524215698,
      "eval_runtime": 6.7625,
      "eval_samples_per_second": 554.532,
      "eval_steps_per_second": 8.725,
      "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
}