基于pytorch123的pytorch入门教程（1）

基于pytorch123的pytorch入门教程（1）

import torch

创建tensor

#创建tensor
x = torch.empty(5,3)
print(x)

tensor([[9.2755e-39, 1.0837e-38, 8.4490e-39],
        [1.1112e-38, 1.0194e-38, 9.0919e-39],
        [8.4490e-39, 9.6429e-39, 8.4490e-39],
        [9.6429e-39, 9.2755e-39, 1.0286e-38],
        [9.0919e-39, 8.9082e-39, 9.2755e-39]])

#创建随机tensor
x = torch.rand(5,3)
print(x)

tensor([[0.6126, 0.4628, 0.1457],
        [0.9288, 0.1584, 0.1483],
        [0.8208, 0.2298, 0.5930],
        [0.7131, 0.9933, 0.5642],
        [0.2639, 0.2375, 0.1098]])

#创建long型tensor
x = torch.zeros(5,3,dtype=torch.long)
print(x)
x.dtype

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





torch.int64

x = x.new_ones(5,3)
x.dtype
#与上一cell的x拥有相同的dtype

torch.int64

tensor运算

#直接创建，叶子节点
x = torch.ones(2,2,requires_grad=True)
print(x)
print(x.grad_fn)

tensor([[1., 1.],
        [1., 1.]], requires_grad=True)
None

#加法创建y，拥有grad_fn
y = x + 2
print(y)
print(y.grad_fn)

tensor([[3., 3.],
        [3., 3.]], grad_fn=)

print(x.is_leaf,y.is_leaf)
#x是叶子节点，y不是叶子节点

True False

#来点复杂节点
z = y * y * 3
out = z.mean()
print(z,out)

tensor([[27., 27.],
        [27., 27.]], grad_fn=) tensor(27., grad_fn=)

输出z与z的平均值，grad_fn有响应的值

#requires_grad属性改变
a = torch.randn(2,2)
a = ((a*3)/(a-1))
print(a.requires_grad)#false
a.requires_grad_(True)
print(a.requires_grad)#true
b = (a*a).sum()
print(b.grad_fn)

False
True

梯度

#out反向求导
out.backward()

F:Anaconda3envspytorchlibsite-packagestorchautograd__init__.py:149: UserWarning: CUDA initialization: The NVIDIA driver on your system is too old (found version 9020). Please update your GPU driver by downloading and installing a new version from the URL: http://www.nvidia.com/Download/index.aspx Alternatively, go to: https://pytorch.org to install a PyTorch version that has been compiled with your version of the CUDA driver. (Triggered internally at  ..c10cudaCUDAFunctions.cpp:115.)
  allow_unreachable=True, accumulate_grad=True)  # allow_unreachable flag

#out对x的导数
print(x.grad)

tensor([[4.5000, 4.5000],
        [4.5000, 4.5000]])

x

tensor([[1., 1.],
        [1., 1.]], requires_grad=True)

out2 = x.sum()
out2.backward()
print(x.grad)
#注意grad是累加的

tensor([[5.5000, 5.5000],
        [5.5000, 5.5000]])

out3 = x.sum()
x.grad.data.zero_() #x梯度清零
out3.backward()
print(x.grad)

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