trainer_state.json

{
  "best_metric": 0.888,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0747/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": 0.0002886495269365962,
      "loss": 2.1606,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6714666666666667,
      "eval_loss": 1.1270029544830322,
      "eval_runtime": 7.9458,
      "eval_samples_per_second": 471.949,
      "eval_steps_per_second": 7.425,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.768027305603027,
      "learning_rate": 0.00025619102456015866,
      "loss": 0.756,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7637333333333334,
      "eval_loss": 0.7851364612579346,
      "eval_runtime": 7.04,
      "eval_samples_per_second": 532.671,
      "eval_steps_per_second": 8.381,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0002077292077256867,
      "loss": 0.4199,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8088,
      "eval_loss": 0.6285783648490906,
      "eval_runtime": 6.8792,
      "eval_samples_per_second": 545.118,
      "eval_steps_per_second": 8.577,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.074754476547241,
      "learning_rate": 0.00015035378262848306,
      "loss": 0.219,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8434666666666667,
      "eval_loss": 0.5491188764572144,
      "eval_runtime": 7.0444,
      "eval_samples_per_second": 532.335,
      "eval_steps_per_second": 8.375,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.24327822029590607,
      "learning_rate": 9.29244973331404e-05,
      "loss": 0.1007,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8570666666666666,
      "eval_loss": 0.5100055932998657,
      "eval_runtime": 6.8872,
      "eval_samples_per_second": 544.485,
      "eval_steps_per_second": 8.567,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.09456682950258255,
      "learning_rate": 4.4184439928107014e-05,
      "loss": 0.0298,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.88,
      "eval_loss": 0.4572104811668396,
      "eval_runtime": 6.8641,
      "eval_samples_per_second": 546.321,
      "eval_steps_per_second": 8.595,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.19480907917022705,
      "learning_rate": 1.1553842323578666e-05,
      "loss": 0.0086,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.888,
      "eval_loss": 0.4407956004142761,
      "eval_runtime": 6.8709,
      "eval_samples_per_second": 545.778,
      "eval_steps_per_second": 8.587,
      "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
}