3-23Pytorch与张量切片

最新推荐文章于 2024-02-24 23:45:00 发布

范德彪陕西分彪

最新推荐文章于 2024-02-24 23:45:00 发布

阅读量248

点赞数

分类专栏： pytorch学习

本文链接：https://blog.csdn.net/weixin_46815330/article/details/113099190

版权

pytorch学习专栏收录该内容

138 篇文章 12 订阅

订阅专栏

在这里插入图片描述

import torch
a = torch.rand((10,4))
print(a)
out = torch.chunk(a,2,dim = 0)
print(out[0],out[0].shape)
print(out[1],out[1].shape)

tensor([[0.4780, 0.9652, 0.8689, 0.6855],
        [0.6050, 0.6120, 0.8252, 0.4419],
        [0.8837, 0.2002, 0.0820, 0.7129],
        [0.7875, 0.6649, 0.1209, 0.1815],
        [0.5611, 0.5992, 0.1595, 0.8397],
        [0.3782, 0.1737, 0.2253, 0.7230],
        [0.5463, 0.1684, 0.7635, 0.9761],
        [0.6934, 0.2580, 0.9594, 0.8153],
        [0.9524, 0.3471, 0.6847, 0.8776],
        [0.4589, 0.3072, 0.7128, 0.6409]])
tensor([[0.4780, 0.9652, 0.8689, 0.6855],
        [0.6050, 0.6120, 0.8252, 0.4419],
        [0.8837, 0.2002, 0.0820, 0.7129],
        [0.7875, 0.6649, 0.1209, 0.1815],
        [0.5611, 0.5992, 0.1595, 0.8397]]) torch.Size([5, 4])
tensor([[0.3782, 0.1737, 0.2253, 0.7230],
        [0.5463, 0.1684, 0.7635, 0.9761],
        [0.6934, 0.2580, 0.9594, 0.8153],
        [0.9524, 0.3471, 0.6847, 0.8776],
        [0.4589, 0.3072, 0.7128, 0.6409]]) torch.Size([5, 4])

out = torch.chunk(a,2,dim = 1)
print(out[0],out[0].shape)
print(out[1],out[1].shape)


tensor([[0.4780, 0.9652],
        [0.6050, 0.6120],
        [0.8837, 0.2002],
        [0.7875, 0.6649],
        [0.5611, 0.5992],
        [0.3782, 0.1737],
        [0.5463, 0.1684],
        [0.6934, 0.2580],
        [0.9524, 0.3471],
        [0.4589, 0.3072]]) torch.Size([10, 2])
tensor([[0.8689, 0.6855],
        [0.8252, 0.4419],
        [0.0820, 0.7129],
        [0.1209, 0.1815],
        [0.1595, 0.8397],
        [0.2253, 0.7230],
        [0.7635, 0.9761],
        [0.9594, 0.8153],
        [0.6847, 0.8776],
        [0.7128, 0.6409]]) torch.Size([10, 2])

out = torch.split(a,2,dim=0)
print(out)

(tensor([[0.4780, 0.9652, 0.8689, 0.6855],
        [0.6050, 0.6120, 0.8252, 0.4419]]), tensor([[0.8837, 0.2002, 0.0820, 0.7129],
        [0.7875, 0.6649, 0.1209, 0.1815]]), tensor([[0.5611, 0.5992, 0.1595, 0.8397],
        [0.3782, 0.1737, 0.2253, 0.7230]]), tensor([[0.5463, 0.1684, 0.7635, 0.9761],
        [0.6934, 0.2580, 0.9594, 0.8153]]), tensor([[0.9524, 0.3471, 0.6847, 0.8776],
        [0.4589, 0.3072, 0.7128, 0.6409]]))

out = torch.split(a,3,dim=0)
print(out)
print(len(out))

(tensor([[0.4780, 0.9652, 0.8689, 0.6855],
        [0.6050, 0.6120, 0.8252, 0.4419],
        [0.8837, 0.2002, 0.0820, 0.7129]]), tensor([[0.7875, 0.6649, 0.1209, 0.1815],
        [0.5611, 0.5992, 0.1595, 0.8397],
        [0.3782, 0.1737, 0.2253, 0.7230]]), tensor([[0.5463, 0.1684, 0.7635, 0.9761],
        [0.6934, 0.2580, 0.9594, 0.8153],
        [0.9524, 0.3471, 0.6847, 0.8776]]), tensor([[0.4589, 0.3072, 0.7128, 0.6409]]))
4
tensor([[0.4780, 0.9652, 0.8689, 0.6855],
        [0.6050, 0.6120, 0.8252, 0.4419],
        [0.8837, 0.2002, 0.0820, 0.7129]]) torch.Size([3, 4])
tensor([[0.7875, 0.6649, 0.1209, 0.1815],
        [0.5611, 0.5992, 0.1595, 0.8397],
        [0.3782, 0.1737, 0.2253, 0.7230]]) torch.Size([3, 4])
tensor([[0.5463, 0.1684, 0.7635, 0.9761],
        [0.6934, 0.2580, 0.9594, 0.8153],
        [0.9524, 0.3471, 0.6847, 0.8776]]) torch.Size([3, 4])
tensor([[0.4589, 0.3072, 0.7128, 0.6409]]) torch.Size([1, 4])

out = torch.split(a,[1,3,6],dim=0)
for t in out:
    print(t,t.shape)

tensor([[0.4780, 0.9652, 0.8689, 0.6855]]) torch.Size([1, 4])
tensor([[0.6050, 0.6120, 0.8252, 0.4419],
        [0.8837, 0.2002, 0.0820, 0.7129],
        [0.7875, 0.6649, 0.1209, 0.1815]]) torch.Size([3, 4])
tensor([[0.5611, 0.5992, 0.1595, 0.8397],
        [0.3782, 0.1737, 0.2253, 0.7230],
        [0.5463, 0.1684, 0.7635, 0.9761],
        [0.6934, 0.2580, 0.9594, 0.8153],
        [0.9524, 0.3471, 0.6847, 0.8776],
        [0.4589, 0.3072, 0.7128, 0.6409]]) torch.Size([6, 4])

范德彪陕西分彪

关注

0
点赞
踩
0

收藏

觉得还不错? 一键收藏
0
评论
3-23Pytorch与张量切片

import torcha = torch.rand((10,4))print(a)out = torch.chunk(a,2,dim = 0)print(out[0],out[0].shape)print(out[1],out[1].shape)tensor([[0.4780, 0.9652, 0.8689, 0.6855], [0.6050, 0.6120, 0.8252, 0.4419], [0.8837, 0.2002, 0.0820, 0.7129].
复制链接

扫一扫

专栏目录