1.官網 https://pytorch.org/get-started/locally/
2.創建pytorch1.5.1的虛擬環境,python版本3.6
conda create -n pytorch1.5.1 python=3.6
3.根據官網說明生成安裝命令,安裝不需要cuda支持的pytorch1.5.1 cpu版:
conda install pytorch torchvision cpuonly -c pytorch
4.測試安裝結果
import torch
print(torch.__version__)
輸出爲‘1.5.1’,表明安裝成功。
5.一個最小的測試例子a.py
# -*- coding: utf-8 -*-
import torch
dtype = torch.float
device = torch.device("cpu")
# device = torch.device("cuda:0") # Uncomment this to run on GPU
# N is batch size; D_in is input dimension;
# H is hidden dimension; D_out is output dimension.
N, D_in, H, D_out = 64, 1000, 100, 10
# Create random input and output data
x = torch.randn(N, D_in, device=device, dtype=dtype)
y = torch.randn(N, D_out, device=device, dtype=dtype)
# Randomly initialize weights
w1 = torch.randn(D_in, H, device=device, dtype=dtype)
w2 = torch.randn(H, D_out, device=device, dtype=dtype)
learning_rate = 1e-6
for t in range(500):
# Forward pass: compute predicted y
h = x.mm(w1)
h_relu = h.clamp(min=0)
y_pred = h_relu.mm(w2)
# Compute and print loss
loss = (y_pred - y).pow(2).sum().item()
print(t, loss)
# Backprop to compute gradients of w1 and w2 with respect to loss
grad_y_pred = 2.0 * (y_pred - y)
grad_w2 = h_relu.t().mm(grad_y_pred)
grad_h_relu = grad_y_pred.mm(w2.t())
grad_h = grad_h_relu.clone()
grad_h[h < 0] = 0
grad_w1 = x.t().mm(grad_h)
# Update weights using gradient descent
w1 -= learning_rate * grad_w1
w2 -= learning_rate * grad_w2
輸出結果:
0 29138108.0
1 23484640.0
2 20555388.0
...
496 3.662774179247208e-05
497 3.607522012316622e-05
498 3.569219552446157e-05
499 3.529985042405315e-05