trainer_state.json

{
  "best_metric": 0.9581333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0735/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": 1.5538221597671509,
      "learning_rate": 9.698463103929542e-05,
      "loss": 0.6019,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9309333333333333,
      "eval_loss": 0.24768637120723724,
      "eval_runtime": 4.5619,
      "eval_samples_per_second": 822.019,
      "eval_steps_per_second": 12.933,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2939869165420532,
      "learning_rate": 8.83022221559489e-05,
      "loss": 0.1431,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9333333333333333,
      "eval_loss": 0.22357867658138275,
      "eval_runtime": 4.6245,
      "eval_samples_per_second": 810.904,
      "eval_steps_per_second": 12.758,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.2912994921207428,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.0751,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9426666666666667,
      "eval_loss": 0.22737278044223785,
      "eval_runtime": 4.6525,
      "eval_samples_per_second": 806.027,
      "eval_steps_per_second": 12.681,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.013050064444541931,
      "learning_rate": 5.868240888334653e-05,
      "loss": 0.0381,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9437333333333333,
      "eval_loss": 0.21776233613491058,
      "eval_runtime": 4.7916,
      "eval_samples_per_second": 782.627,
      "eval_steps_per_second": 12.313,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.03426516056060791,
      "learning_rate": 4.131759111665349e-05,
      "loss": 0.0169,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9496,
      "eval_loss": 0.2301751971244812,
      "eval_runtime": 4.6596,
      "eval_samples_per_second": 804.785,
      "eval_steps_per_second": 12.662,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.005051807966083288,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.0088,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9568,
      "eval_loss": 0.18682152032852173,
      "eval_runtime": 5.2834,
      "eval_samples_per_second": 709.777,
      "eval_steps_per_second": 11.167,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.1731488851999511e-05,
      "loss": 0.0042,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9581333333333333,
      "eval_loss": 0.1894504576921463,
      "eval_runtime": 4.9612,
      "eval_samples_per_second": 755.871,
      "eval_steps_per_second": 11.892,
      "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
}