torch.Tensor.nanmean

张量.nanmean(dim=None, keepdim=False, *, dtype=None) → 张量

参见 torch.nanmean()