Euclidean distance is convenient, but the NumPy axes trick works for arbitrary functions and it works in PyTorch directly (read more about it here ). x ...
確定! 回上一頁