pytorch基础学习一

pytorch基础

 

In [1]:

import torch

 

In [2]:

# 查询torch版本
print(torch.__version__)

 

 

1.11.0+cpu

 

In [3]:

x = torch.rand(5,3)
y = torch.rand(5,3)

 

In [4]:

# 矩阵相加
z = torch.add(x,y)
z

 

Out[4]:

tensor([[0.5727, 0.6683, 1.4929],
        [1.5323, 1.7757, 0.3782],
        [1.5451, 0.4119, 1.0949],
        [0.6613, 0.9304, 1.2568],
        [1.4187, 0.8155, 0.9888]])

 

In [5]:

# 矩阵索引
print(z[1,2],z[:,2],z[1,:])

 

 

tensor(0.3782) tensor([1.4929, 0.3782, 1.0949, 1.2568, 0.9888]) tensor([1.5323, 1.7757, 0.3782])

 

In [6]:

# view矩阵变换
print(z.view([3,5]))

 

 

tensor([[0.5727, 0.6683, 1.4929, 1.5323, 1.7757],
        [0.3782, 1.5451, 0.4119, 1.0949, 0.6613],
        [0.9304, 1.2568, 1.4187, 0.8155, 0.9888]])

 

In [7]:

# 自动计算维度
print(z.view(-1,5))

 

 

tensor([[0.5727, 0.6683, 1.4929, 1.5323, 1.7757],
        [0.3782, 1.5451, 0.4119, 1.0949, 0.6613],
        [0.9304, 1.2568, 1.4187, 0.8155, 0.9888]])