CUDA 11.7 版本下安装Tensorflow-gpu_cuda11.7安装tensorflow

Windows 下 3050显卡，CUDA 11.7 版本，cudnn8.7

安装Tensorflow-gpu

conda create -n tf_gpu_py3.8 python=3.8.0

conda activate tf_gpu_py3.8
pip install tensorflow-gpu==2.6.0

会报错

完整报错如下：

TypeError: Descriptors cannot not be created directly.
If this call came from a _pb2.py file, your generated code is out of date and must be regenerated with protoc >= 3.19.0.
If you cannot immediately regenerate your protos, some other possible workarounds are:
 1. Downgrade the protobuf package to 3.20.x or lower.
 2. Set PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python (but this will use pure-Python parsing and will be much slower).

解决方案：

首先看了网上的大佬说要将protobuf更换成较低的版本

1. pip uninstall protobuf

2. pip install protobuf==3.20.0

即可正确安装

Tensorflow2.6安装什么版本的kears    pip install keras==2.6

numpy版本对应pandas

pip uninstall numpy

pip install numpy==1.19.5

pip install pandas==1.0.4

pip install seaborn==0.9.0

pip install matplotlib==3.3.4

pip install scipy==1.5.4