2021-10-19 PyTorch教程之Tensors(一）

最新推荐文章于 2025-07-16 07:30:18 发布

原创最新推荐文章于 2025-07-16 07:30:18 发布 · 217 阅读

0 ·

CC 4.0 BY-SA版权

文章标签：

#pytorch #深度学习 #神经网络

深度学习专栏收录该内容

57 篇文章

订阅专栏

本文介绍了如何使用PyTorch创建Tensor对象，展示基本的数学运算，如加法、大小获取，以及切片操作。通过实例演示了`torch.Tensor`的创建、元素级操作和维度处理，适合初学者理解Tensors的核心概念。

部署运行你感兴趣的模型镜像

PyTorch教程之Tensors

>>> import torch
>>> x = torch.Tensor(2,3)
>>> x
tensor([[0., 0., 0.],
[0., 0., 0.]])
>>> y=torch.rand(3,4)
>>> y
tensor([[0.4464, 0.8380, 0.2633, 0.6285],
[0.9167, 0.4549, 0.4253, 0.6597],
[0.0015, 0.4274, 0.2199, 0.3969]])
>>> y.size()
torch.Size([3, 4])
>>> a = y.size()[0]
>>> a
3
>>> b = y.size()[1]
>>> b
4
>>> z = torch.rand(3,4)
>>> z
tensor([[0.3165, 0.9564, 0.0270, 0.6598],
[0.6044, 0.8573, 0.3340, 0.8657],
[0.0120, 0.6371, 0.1377, 0.8685]])
>>> y+z
tensor([[0.7629, 1.7944, 0.2903, 1.2883],
[1.5211, 1.3122, 0.7593, 1.5254],
[0.0136, 1.0645, 0.3576, 1.2654]])
>>> torch.add(y,z)
tensor([[0.7629, 1.7944, 0.2903, 1.2883],
[1.5211, 1.3122, 0.7593, 1.5254],
[0.0136, 1.0645, 0.3576, 1.2654]])
>>> y.add_(z)
tensor([[0.7629, 1.7944, 0.2903, 1.2883],
[1.5211, 1.3122, 0.7593, 1.5254],
[0.0136, 1.0645, 0.3576, 1.2654]])
>>> y
tensor([[0.7629, 1.7944, 0.2903, 1.2883],
[1.5211, 1.3122, 0.7593, 1.5254],
[0.0136, 1.0645, 0.3576, 1.2654]])
>>> y[:,1]
tensor([1.7944, 1.3122, 1.0645])
>>> y[1,:]
tensor([1.5211, 1.3122, 0.7593, 1.5254])
>>>