学术加速

受国内网络法规限制，访问github等海外学术网站时，网络速度会时快时慢。为方便大家学习，我们提供了部分站点的学术加速，特声明如下：

声明：

1.HTTP代理地址为：172.168.1.240:4080,使用时请遵守相关法律法规

2.仅开通github.com python.org pypi.org pytorch.org等网站加速

3.学术加速为友情提供，不保证长期性和稳定性

更换ubuntu系统apt源，pip源，conda源脚本

一。HuggingFace模型下载慢

由于国内禁止访问HuggingFace网站，推荐使用国内的镜像站https://hf-mirror.com/,

或联系客服下载到公共网盘

二。安装pytorch和tensorflow速度慢

解决方法：使用华益云内网镜像源，秒级安装

1.pytorch安装

#CUDA11.8

pip install torch==2.3.1 torchvision==0.18.1 torchaudio==2.3.1  --index-url http://172.168.200.101/pytorch/ --trusted-host 172.168.200.101
pip install torch==2.3.0 torchvision==0.18.0 torchaudio==2.3.0  --index-url http://172.168.200.101/pytorch/ --trusted-host 172.168.200.101
pip install torch==2.2.2 torchvision==0.17.2 torchaudio==2.2.2  --index-url http://172.168.200.101/pytorch/ --trusted-host 172.168.200.101
pip install torch==2.2.1 torchvision==0.17.1 torchaudio==2.2.1  --index-url http://172.168.200.101/pytorch/ --trusted-host 172.168.200.101
pip install torch==2.2.0 torchvision==0.17.0 torchaudio==2.2.0  --index-url http://172.168.200.101/pytorch/ --trusted-host 172.168.200.101
pip install torch==2.1.2 torchvision==0.16.2 torchaudio==2.1.2  --index-url http://172.168.200.101/pytorch/ --trusted-host 172.168.200.101
pip install torch==2.1.1 torchvision==0.16.1 torchaudio==2.1.1  --index-url http://172.168.200.101/pytorch/ --trusted-host 172.168.200.101
pip install torch==2.1.0 torchvision==0.16.0 torchaudio==2.1.0  --index-url http://172.168.200.101/pytorch/ --trusted-host 172.168.200.101
pip install torch==2.0.1 torchvision==0.15.2 torchaudio==2.0.2  --index-url http://172.168.200.101/pytorch/ --trusted-host 172.168.200.101
pip install torch==2.0.0 torchvision==0.15.1 torchaudio==2.0.1  --index-url http://172.168.200.101/pytorch/ --trusted-host 172.168.200.101

#CUDA11.1

pip install torch==1.9.1+cu111 torchvision==0.10.1+cu111 torchaudio==0.9.1  --index-url http://172.168.200.101/pytorch/ --trusted-host 172.168.200.101
pip install torch==1.8.1+cu111 torchvision==0.9.1+cu111 torchaudio==0.8.1
pip install torch==1.8.0+cu111 torchvision==0.9.0+cu111 torchaudio==0.8.0

#CUDA10.1

pip install torch==1.7.1+cu101 torchvision==0.8.2+cu101 torchaudio==0.7.2
pip install torch==1.8.0+cu101 torchvision==0.9.0+cu101 torchaudio==0.8.0

2.tensorflow安装

pip install tensorflow-gpu==2.9.0   --index-url http://172.168.200.101/tensorflow/ --trusted-host 172.168.200.101
pip install tensorflow-gpu==2.8.0   --index-url http://172.168.200.101/tensorflow/ --trusted-host 172.168.200.101
pip install tensorflow-gpu==2.7.0   --index-url http://172.168.200.101/tensorflow/ --trusted-host 172.168.200.101
pip install tensorflow-gpu==2.6.0   --index-url http://172.168.200.101/tensorflow/ --trusted-host 172.168.200.101
pip install tensorflow-gpu==2.5.0   --index-url http://172.168.200.101/tensorflow/ --trusted-host 172.168.200.101
pip install tensorflow-gpu==2.4.0   --index-url http://172.168.200.101/tensorflow/ --trusted-host 172.168.200.101
pip install tensorflow-gpu==2.3.0   --index-url http://172.168.200.101/tensorflow/ --trusted-host 172.168.200.101
pip install tensorflow-gpu==2.2.0   --index-url http://172.168.200.101/tensorflow/ --trusted-host 172.168.200.101

三。apt命令下载慢：

解决方法：切换apt镜像源

阿里说明文档：https://developer.aliyun.com/mirror/ubuntu?spm=a2c6h.13651102.0.0.24a31b110HFUi9

参考：https://blog.csdn.net/gumc123/article/details/140186779

vim /etc/apt/sources.list 写入：

ubuntu22.04：jammy

ubuntu20.04：focal

ubuntu18.04：bionic

清华源

deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy main restricted universe multiverse
deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy-updates main restricted universe multiverse
deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy-backports main restricted universe multiverse
deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy-security main restricted universe multiverse

阿里源

deb http://mirrors.aliyun.com/ubuntu/ jammy main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu/ jammy-updates main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu/ jammy-backports main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu/ jammy-security main restricted universe multiverse

官方源

deb http://cn.archive.ubuntu.com/ubuntu/ jammy main restricted
deb http://cn.archive.ubuntu.com/ubuntu/ jammy-updates main restricted
deb http://cn.archive.ubuntu.com/ubuntu/ jammy universe
deb http://cn.archive.ubuntu.com/ubuntu/ jammy-updates universe
deb http://cn.archive.ubuntu.com/ubuntu/ jammy multiverse
deb http://cn.archive.ubuntu.com/ubuntu/ jammy-updates multiverse
deb http://cn.archive.ubuntu.com/ubuntu/ jammy-backports main restricted universe multiverse
deb http://security.ubuntu.com/ubuntu jammy-security main restricted
deb http://security.ubuntu.com/ubuntu jammy-security universe
deb http://security.ubuntu.com/ubuntu jammy-security multiverse

四。pip安装下载速度慢

解决方法：更换源或使用学术加速

1. 更换源

#阿里源
pip install 包名 -i https://mirrors.aliyun.com/pypi/simple/
#清华源
pip install 包名 -i https://pypi.tuna.tsinghua.edu.cn/simple/
#华为源
pip install 包名 -i https://repo.huaweicloud.com/repository/pypi/simple/
#腾讯源
pip install 包名 -i http://mirrors.cloud.tencent.com/pypi/simple

2. 使用学术加速

pip install --proxy=http://172.168.1.240:4080 包名

五。wget下载速度慢

解决方法：查看是否是国外的，如果是国外的，请使用学术加速

export http_proxy=http://172.168.1.240:4080
export https_proxy=http://172.168.1.240:4080

#查看代理是否生效

env | grep -i proxy

#下载资源

wget https://repo.anaconda.com/miniconda/Miniconda3-py38_4.12.0-Linux-x86_64.sh

#关闭代理

unset http_proxy
unset https_proxy

六。wget下载CUDA速度慢

解决方法：换为国内地址

例如：

wget https://developer.download.nvidia.com/compute/cuda/11.8.0/local_installers/cuda_11.8.0_520.61.05_linux.run

请将：developer.download.nvidia.com 更改为：developer.download.nvidia.cn

七。Git clone速度慢

解决方法：学术加速

git clone --config http.sslVerify=false --config https.sslVerify=false --config http.proxy=http://172.168.1.240:4080 --config https.proxy=http://172.168.1.240:4080 https://github.com/Dao-AILab/flash-attention