PyTorch学习笔记

[wang910930]

PyTorch
# 安装NVIDIA cuda
https://developer.nvidia.com/cuda-downloads?target_os=Windows&target_arch=x86_64&target_version=10

# GPU compute capability <= 3.5 无法使用1.3版本及以上的PyTorch
conda install pytorch==1.2.0 torchvision==0.4.0 cudatoolkit=10.0


# 清华镜像
conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/pytorch/
conda config --set show_channel_urls yes

# 下载
conda install pytorch torchvision cudatoolkit=10.1

# 检验运行成功
import torch
import torchvision
a = torch.Tensor(5, 3)
a = a.cuda()
print(a)

# 常用包调用
import torch
import torch.nn.functional as F
import torch.utils.data
from torchvision import datasets, transforms
import torch.optim as optim
import torch.nn as nn