torch.stack()和torch.cat()区别

最新推荐文章于 2024-09-15 18:10:00 发布

Cakker

最新推荐文章于 2024-09-15 18:10:00 发布

阅读量145

点赞数

文章标签：深度学习 python pytorch

本文链接：https://blog.csdn.net/qq_43049542/article/details/125821579

版权

>>> import torch
>>> T1=torch.randn([2,3])
>>> T2=torch.randn([2,3])

>>> T1
tensor([[-0.5566, -1.2475, -0.5865],
        [-2.3513,  0.4147, -1.0349]])
>>> T2
tensor([[-0.1764, -0.8736,  0.9022],
        [-1.1520, -0.1529, -0.1760]])


>>> Tcat=torch.cat([T1,T2],dim=0)
>>> Tcat
tensor([[-0.5566, -1.2475, -0.5865],
        [-2.3513,  0.4147, -1.0349],
        [-0.1764, -0.8736,  0.9022],
        [-1.1520, -0.1529, -0.1760]])
>>> Tcat.shape
torch.Size([4, 3])


>>> Tstack=torch.stack([T1,T2],dim=0)
>>> Tstack
tensor([[[-0.5566, -1.2475, -0.5865],
         [-2.3513,  0.4147, -1.0349]],

        [[-0.1764, -0.8736,  0.9022],
         [-1.1520, -0.1529, -0.1760]]])
>>> Tstack.shape
torch.Size([2, 2, 3])