PyTorch 60 分钟入门教程:PyTorch 深度学习官方入门中文教程
什么是 PyTorch?
PyTorch 是一个基于 Python 的科学计算包,主要定位两类人群:- NumPy 的替代品,可以利用 GPU 的性能进行计算。
- 深度学习研究平台拥有足够的灵活性和速度
开始学习
Tensors (张量)
Tensors 类似于 NumPy 的 ndarrays ,同时 Tensors 可以使用 GPU 进行计算。from __future__ import print_function
import torch
x = torch.empty(5, 3)
print(x)
输出:
tensor(1.00000e-04 *
[[-0.0000, 0.0000, 1.5135],
[ 0.0000, 0.0000, 0.0000],
[ 0.0000, 0.0000, 0.0000],
[ 0.0000, 0.0000, 0.0000],
[ 0.0000, 0.0000, 0.0000]])
构造一个随机初始化的矩阵:
x = torch.rand(5, 3)
print(x)
输出:
tensor([[ 0.6291, 0.2581, 0.6414],
[ 0.9739, 0.8243, 0.2276],
[ 0.4184, 0.1815, 0.5131],
[ 0.5533, 0.5440, 0.0718],
[ 0.2908, 0.1850, 0.5297]])
构造一个矩阵全为 0,而且数据类型是 long.
Construct a matrix filled zeros and of dtype long:
x = torch.zeros(5, 3, dtype=torch.long)
print(x)
输出:
tensor([[ 0, 0, 0],
[ 0, 0, 0],
[ 0, 0, 0],
[ 0, 0, 0],
[ 0, 0, 0]])
x = torch.tensor([5.5, 3])
print(x)
输出:
tensor([ 5.5000, 3.0000])
x = x.new_ones(5, 3, dtype=torch.double)
# new_* methods take in sizes
print(x)
x = torch.randn_like(x, dtype=torch.float)
# override dtype!
print(x)
# result has the same size
输出:
tensor([[ 1., 1., 1.],
[ 1., 1., 1.],
[ 1., 1., 1.],
[ 1., 1., 1.],
[ 1., 1., 1.]], dtype=torch.float64)
tensor([[-0.2183, 0.4477, -0.4053],
[ 1.7353, -0.0048, 1.2177],
[-1.1111, 1.0878, 0.9722],
[-0.7771, -0.2174, 0.0412],
[-2.1750, 1.3609, -0.3322]])
print(x.size())
输出:
torch.Size([5, 3])
注意
torch.Size
是一个元组,所以它支持左右的元组操作。
操作
在接下来的例子中,我们将会看到加法操作。加法: 方式 1
y = torch.rand(5, 3)
print(x + y)
Out:
tensor([[-0.1859, 1.3970, 0.5236],
[ 2.3854, 0.0707, 2.1970],
[-0.3587, 1.2359, 1.8951],
[-0.1189, -0.1376, 0.4647],
[-1.8968, 2.0164, 0.1092]])
print(torch.add(x, y))
Out:
tensor([[-0.1859, 1.3970, 0.5236],
[ 2.3854, 0.0707, 2.1970],
[-0.3587, 1.2359, 1.8951],
[-0.1189, -0.1376, 0.4647],
[-1.8968, 2.0164, 0.1092]])
result = torch.empty(5, 3)
torch.add(x, y, out=result)
print(result)
Out:
tensor([[-0.1859, 1.3970, 0.5236],
[ 2.3854, 0.0707, 2.1970],
[-0.3587, 1.2359, 1.8951],
[-0.1189, -0.1376, 0.4647],
[-1.8968, 2.0164, 0.1092]])
# adds x to y
y.add_(x)
print(y)
Out:
tensor([[-0.1859, 1.3970, 0.5236],
[ 2.3854, 0.0707, 2.1970],
[-0.3587, 1.2359, 1.8951],
[-0.1189, -0.1376, 0.4647],
[-1.8968, 2.0164, 0.1092]])
Note
注意任何使张量会发生变化的操作都有一个前缀 ''。例如:x.copy
(y)
, x.t_()
, 将会改变 x
.
你可以使用标准的 NumPy 类似的索引操作
print(x[:, 1])
Out:
tensor([ 0.4477, -0.0048, 1.0878, -0.2174, 1.3609])
torch.view
:
x = torch.randn(4, 4)
y = x.view(16)
z = x.view(-1, 8) # the size -1 is inferred from other dimensions
print(x.size(), y.size(), z.size())
Out:
torch.Size([4, 4]) torch.Size([16]) torch.Size([2, 8])
x = torch.randn(1)
print(x)
print(x.item())
Out:
tensor([ 0.9422])
0.9422121644020081