谢谢大佬的开源项目，运行的很成功。

[6025jie]

运行环境：
python36
Anaconda3
opencv-python==4.3.0.38
tensorflow==1.5
scikit-learn==0.22.2

[lohithnh]

@6025jie Could you please explain how did you set it up ?
I'm trying on macbook

[6025jie]

centos7

sudo rpm --import https://dl.fedoraproject.org/pub/epel/RPM-GPG-KEY-EPEL-7

sudo yum install epel-release
sudo yum install centos-release-scl

安装 Python 3.6

sudo yum install rh-python36

启用 Python 3.6

scl enable rh-python36 bash

Anaconda安装

# 清华源
https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/

wget https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/Anaconda3-2023.03-1-Linux-x86_64.sh

#  验证下载的文件（可选）
sha256sum Anaconda3-2023.03-1-Linux-x86_64.sh

# 运行安装脚本
bash Anaconda3-2023.03-1-Linux-x86_64.sh

# 初始化 Anaconda
source ~/anaconda3/bin/activate

# 更新 Anaconda
conda update conda
conda update anaconda

# 验证安装
conda --version

conda init

conda create -n python36 python=3.6

conda list

conda activate python36

pip install --upgrade pip

pip install opencv-python==4.3.0.38 -i https://pypi.tuna.tsinghua.edu.cn/simple

pip uninstall tensorflow

pip install tensorflow-gpu==1.5 -i https://pypi.tuna.tsinghua.edu.cn/simple
或者
pip install tensorflow==1.5 -i https://pypi.tuna.tsinghua.edu.cn/simple

pip install scikit-learn==0.22.2 -i https://pypi.tuna.tsinghua.edu.cn/simple

pip install numba -i https://pypi.tuna.tsinghua.edu.cn/simple
pip install numpy -i https://pypi.tuna.tsinghua.edu.cn/simple
pip install filterpy -i https://pypi.tuna.tsinghua.edu.cn/simple

sudo yum install mesa-libGL

CPU 不支持 AVX/AVX2，建议你：
安装不使用 AVX 的 TensorFlow 版本：
Tensorflow从1.6开始从AVX编译二进制文件，所以如果你的CPU不支持AVX
安装旧版（1.5或之前版本）：
pip install tensorflow==1.5 -i https://pypi.tuna.tsinghua.edu.cn/simple

ln -s /usr/local/lib/librocketmq.so /usr/lib
sudo ldconfig