trainer_state.json

{
  "best_metric": 0.9565333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0613/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": 0.4851626455783844,
      "learning_rate": 6.222222222222222e-05,
      "loss": 0.701,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9426666666666667,
      "eval_loss": 0.21022385358810425,
      "eval_runtime": 7.4882,
      "eval_samples_per_second": 500.785,
      "eval_steps_per_second": 7.879,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7092217803001404,
      "learning_rate": 5.444444444444444e-05,
      "loss": 0.1051,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9530666666666666,
      "eval_loss": 0.1648063212633133,
      "eval_runtime": 6.1687,
      "eval_samples_per_second": 607.906,
      "eval_steps_per_second": 9.564,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.10445474088191986,
      "learning_rate": 4.666666666666666e-05,
      "loss": 0.0331,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9493333333333334,
      "eval_loss": 0.18577507138252258,
      "eval_runtime": 6.0789,
      "eval_samples_per_second": 616.884,
      "eval_steps_per_second": 9.706,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.018997058272361755,
      "learning_rate": 3.8888888888888884e-05,
      "loss": 0.0134,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9541333333333334,
      "eval_loss": 0.18360471725463867,
      "eval_runtime": 6.1359,
      "eval_samples_per_second": 611.162,
      "eval_steps_per_second": 9.616,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.014172249473631382,
      "learning_rate": 3.111111111111111e-05,
      "loss": 0.0065,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9530666666666666,
      "eval_loss": 0.19360390305519104,
      "eval_runtime": 6.1392,
      "eval_samples_per_second": 610.825,
      "eval_steps_per_second": 9.61,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.01686074770987034,
      "learning_rate": 2.333333333333333e-05,
      "loss": 0.0029,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.956,
      "eval_loss": 0.19179819524288177,
      "eval_runtime": 6.1372,
      "eval_samples_per_second": 611.026,
      "eval_steps_per_second": 9.613,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.557891224557891e-05,
      "loss": 0.0013,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9565333333333333,
      "eval_loss": 0.1875389963388443,
      "eval_runtime": 6.0854,
      "eval_samples_per_second": 616.224,
      "eval_steps_per_second": 9.695,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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
}