版权声明:本文为博主原创文章,未经博主允许不得转载。 https://blog.csdn.net/jacke121/article/details/83044674
permute
prediction = input.view(bs, self.num_anchors,
self.bbox_attrs, in_h, in_w).permute(0, 1, 3, 4, 2).contiguous()
转置:
import torch x = torch.linspace(1, 9, steps=9).view(3, 3) b=x.permute(1,0) print(b) print(b.permute(1,0))