Pytorch-get ready with me

Posted 2021-01-07 btschang

 

from __future__ import print_function
import torch

x = torch.empty(5, 3) 

print(x)

tensor([[ 3.7740e+04, 4.5877e-41, -5.4795e-33], [ 3.0792e-41, 0.0000e+00, 0.0000e+00], [ 0.0000e+00, 0.0000e+00, 3.4438e-41], [ 0.0000e+00, 4.8901e-36, 2.8026e-45], [ 3.1212e+29, 0.0000e+00, 4.6243e-44]])

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

tensor([[0.1607, 0.0298, 0.7555], [0.8887, 0.1625, 0.6643], [0.7328, 0.5419, 0.6686], [0.0793, 0.1133, 0.5956], [0.3149, 0.9995, 0.6372]])

x = torch.zeros(5, 3, dtype=torch.long)
print(x)

tensor([[0, 0, 0], [0, 0, 0], [0, 0, 0], [0, 0, 0], [0, 0, 0]])

x = torch.tensor([5.5, 3])
print(x)

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

tensor([[1., 1., 1.], [1., 1., 1.], [1., 1., 1.], [1., 1., 1.], [1., 1., 1.]], dtype=torch.float64)

tensor([[-0.2217, -0.9135, -0.6010], [-0.3193, -0.3675, 0.1951], [ 0.0646, -0.4947, 1.0374], [-0.4154, -1.0247, -1.2872], [ 0.5228, 0.3420, 0.0219]])

print(x.size())

torch.Size([5, 3])

【注意】torch.Size 是一个 tuple , 支持所有的tuple操作

 

 

相加

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

tensor([[ 0.2349, -0.0427, -0.5053], [ 0.6455, 0.1199, 0.4239], [ 0.1279, 0.1105, 1.4637], [ 0.4259, -0.0763, -0.9671], [ 0.6856, 0.5047, 0.4250]])

print(torch.add(x, y))

tensor([[ 0.2349, -0.0427, -0.5053], [ 0.6455, 0.1199, 0.4239], [ 0.1279, 0.1105, 1.4637], [ 0.4259, -0.0763, -0.9671], [ 0.6856, 0.5047, 0.4250]])

 

也可以输出给一个tensor

result = torch.empty(5, 3)
torch.add(x, y, out=result)
print(result)

tensor([[ 0.2349, -0.0427, -0.5053], [ 0.6455, 0.1199, 0.4239], [ 0.1279, 0.1105, 1.4637], [ 0.4259, -0.0763, -0.9671], [ 0.6856, 0.5047, 0.4250]])

 

# adds x to y
y.add_(x)
print(y)

tensor([[ 0.2349, -0.0427, -0.5053], [ 0.6455, 0.1199, 0.4239], [ 0.1279, 0.1105, 1.4637], [ 0.4259, -0.0763, -0.9671], [ 0.6856, 0.5047, 0.4250]])

【注意】Any operation that mutates a tensor in-place is post-fixed with an _. For example: x.copy_(y), x.t_(), will change x.

 

输出 x 第二列的所有值

print(x[:, 1])

tensor([-0.9135, -0.3675, -0.4947, -1.0247, 0.3420])

 

用 torch.view resize或reshape tensor

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())

torch.Size([4, 4]) torch.Size([16]) torch.Size([2, 8])

 

用 .item() 提取 tensor存储的数值

x = torch.randn(1)
print(x)
print(x.item())

tensor([1.9218])

1.9218417406082153

 

Tensor 转换成 NumPy Array

a = torch.ones(5)
print(a)

tensor([1., 1., 1., 1., 1.])

b = a.numpy()
print(b)

[1. 1. 1. 1. 1.]

a.add_(1)
print(a)
print(b)

tensor([2., 2., 2., 2., 2.])

[2. 2. 2. 2. 2.]

 

NumPy Array 转换成 Tensor

import numpy as np
a = np.ones(5)
b = torch.from_numpy(a)
np.add(a, 1, out=a)
print(a)
print(b)

[2. 2. 2. 2. 2.]

tensor([2., 2., 2., 2., 2.], dtype=torch.float64)

 

CUDA TENSORS

用  .to 方法

# let us run this cell only if CUDA is available
# We will use ``torch.device`` objects to move tensors in and out of GPU
if torch.cuda.is_available():
    device = torch.device("cuda")          # a CUDA device object
    y = torch.ones_like(x, device=device)  # directly create a tensor on GPU
    x = x.to(device)                       # or just use strings ``.to("cuda")``
    z = x + y
    print(z)
    print(z.to("cpu", torch.double))       # ``.to`` can also change dtype together!

tensor([2.9218], device=‘cuda:0‘)

tensor([2.9218], dtype=torch.float64)