一、安装环境配置
创建、进入虚拟环境，
bash
(base) C:\Users\fuhang>conda create -n py3.8-tf-gpu-2.10.0 python=3.8
Retrieving notices: ...working... done
Channels:
- defaults
Platform: win-64
Collecting package metadata (repodata.json): done
Solving environment: done
==> WARNING: A newer version of conda exists. <==
current version: 23.11.0
latest version: 25.9.0
Please update conda by running
$ conda update -n base -c defaults conda
## Package Plan ##
environment location: D:\ProgramData\Anaconda\envs\py3.8-tf-gpu-2.10.0
added / updated specs:
- python=3.8
The following NEW packages will be INSTALLED:
ca-certificates anaconda/pkgs/main/win-64::ca-certificates-2025.9.9-haa95532_0
libffi anaconda/pkgs/main/win-64::libffi-3.4.4-hd77b12b_1
openssl anaconda/pkgs/main/win-64::openssl-3.0.18-h543e019_0
pip anaconda/pkgs/main/win-64::pip-24.2-py38haa95532_0
python anaconda/pkgs/main/win-64::python-3.8.20-h8205438_0
setuptools anaconda/pkgs/main/win-64::setuptools-75.1.0-py38haa95532_0
sqlite anaconda/pkgs/main/win-64::sqlite-3.50.2-hda9a48d_1
ucrt anaconda/pkgs/main/win-64::ucrt-10.0.22621.0-haa95532_0
vc anaconda/pkgs/main/win-64::vc-14.3-h2df5915_10
vc14_runtime anaconda/pkgs/main/win-64::vc14_runtime-14.44.35208-h4927774_10
vs2015_runtime anaconda/pkgs/main/win-64::vs2015_runtime-14.44.35208-ha6b5a95_10
wheel anaconda/pkgs/main/win-64::wheel-0.44.0-py38haa95532_0二、安装TensorFlow-GPU 2.5.0
pip install tensorflow-gpu==2.5.0 -i https://mirrors.aliyun.com/pypi/simple/
conda list
bash
(py3.8-tf-gpu-2.10.0) C:\Users\fuhang>conda list
# packages in environment at D:\ProgramData\Anaconda\envs\py3.8-tf-gpu-2.10.0:
#
# Name Version Build Channel
absl-py 2.3.1 pypi_0 pypi
astunparse 1.6.3 pypi_0 pypi
ca-certificates 2025.9.9 haa95532_0 defaults
cachetools 5.5.2 pypi_0 pypi
certifi 2025.10.5 pypi_0 pypi
charset-normalizer 3.4.3 pypi_0 pypi
flatbuffers 25.9.23 pypi_0 pypi
gast 0.4.0 pypi_0 pypi
google-auth 2.41.1 pypi_0 pypi
google-auth-oauthlib 0.4.6 pypi_0 pypi
google-pasta 0.2.0 pypi_0 pypi
grpcio 1.70.0 pypi_0 pypi
h5py 3.11.0 pypi_0 pypi
idna 3.10 pypi_0 pypi
importlib-metadata 8.5.0 pypi_0 pypi
keras 2.10.0 pypi_0 pypi
keras-preprocessing 1.1.2 pypi_0 pypi
libclang 18.1.1 pypi_0 pypi
libffi 3.4.4 hd77b12b_1 defaults
markdown 3.7 pypi_0 pypi
markupsafe 2.1.5 pypi_0 pypi
numpy 1.24.4 pypi_0 pypi
oauthlib 3.3.1 pypi_0 pypi
openssl 3.0.18 h543e019_0 defaults
opt-einsum 3.4.0 pypi_0 pypi
packaging 25.0 pypi_0 pypi
pip 24.2 py38haa95532_0 defaults
protobuf 3.19.6 pypi_0 pypi
pyasn1 0.6.1 pypi_0 pypi
pyasn1-modules 0.4.2 pypi_0 pypi
python 3.8.20 h8205438_0 defaults
requests 2.32.4 pypi_0 pypi
requests-oauthlib 2.0.0 pypi_0 pypi
rsa 4.9.1 pypi_0 pypi
setuptools 75.1.0 py38haa95532_0 defaults
six 1.17.0 pypi_0 pypi
sqlite 3.50.2 hda9a48d_1 defaults
tensorboard 2.10.1 pypi_0 pypi
tensorboard-data-server 0.6.1 pypi_0 pypi
tensorboard-plugin-wit 1.8.1 pypi_0 pypi
tensorflow-estimator 2.10.0 pypi_0 pypi
tensorflow-gpu 2.10.0 pypi_0 pypi
tensorflow-io-gcs-filesystem 0.31.0 pypi_0 pypi
termcolor 2.4.0 pypi_0 pypi
typing-extensions 4.13.2 pypi_0 pypi
ucrt 10.0.22621.0 haa95532_0 defaults
urllib3 2.2.3 pypi_0 pypi
vc 14.3 h2df5915_10 defaults
vc14_runtime 14.44.35208 h4927774_10 defaults
vs2015_runtime 14.44.35208 ha6b5a95_10 defaults
werkzeug 3.0.6 pypi_0 pypi
wheel 0.44.0 py38haa95532_0 defaults
wrapt 1.17.3 pypi_0 pypi
zipp 3.20.2 pypi_0 pypi三、查看TensorFlow-GPU 2.5.0是否可用
对于tensorflow - gpu与cuda12.0的相关信息：
兼容性：参考材料中未直接提及tensorflow - gpu与cuda12.0的对应关系。不过可知TensorFlow 2.10.0是Windows本地支持GPU的最后一个版本，从2.11.0开始，Windows系统需通过WSL2才能使用GPU，且GPU和CPU版本在2.x中已合并，安装tensorflow==2.10.0后，若环境配置正确，会自动启用GPU 2。同时，对于Tensorflow2.0，要求CUDA的版本不低于10.0