Volmodaoist·Emanon

It seems like the 'success' value in the return of the 'attack' function is overconfident.

2

It seems like the 'success' value in the return of the 'attack' function is overconfident. ```python if __name__ == '__main__': args = parser.parse_args() model = ModelController(args).get_model().eval() device = torch.device('cuda' if...

Error caused by pytoch version

1

The following code in `build. py` led to an error：torch.utils.ffi is deprecated. Please use cpp extensions instead. ```Python from torch.utils.ffi import create_extension ``` After I replace `from torch.utils.ffi import create_extension`...

对于张量运算应尽量避免使用循环，原版的 model_norm 函数 (i.e. 计算两个模型之间的欧式距离) 使用 for 实现，导致效率偏低且可读性差。为此我们优化了原版 model_norm 方法，增强了可读性且运行效率提高了一倍，若在 GPU 环境之中运行代码， 性能提升会更加显著。 ```python # 原本的 model_norm 方法之中的 for 循环导致其无法充分利用 GPU 加速 def model_norm(model_1, model_2): squared_sum = 0 for name,...