深度学习框架安装日志

1. Download Anaconda:

• Link: Anaconda

2. Download Visual Studio C++ Redistributables:

• Link: Visual Studio C++ Redistributables

3. TensorFlow-GPU Installation:

• Visit TensorFlow Installation Guide and go to the TensorFlow-GPU section.
• Check for the latest versions and requirements.

4. Anaconda Environment Setup:

a. Open Anaconda prompt as administrator.

b. Create a new environment:
conda create -n tf_gpu python==3.9

c. Activate the environment:
conda activate tf_gpu

5. Install CUDA and cuDNN:

a. Check the required versions from the TensorFlow-GPU section (Step 3).

b. Install CUDA and cuDNN using conda:

conda install cudatoolkit=11.2 cudnn=8.1 -c=conda-forge

//国内安装会失败 手工下载地址

coda：https://developer.nvidia.com/cuda-toolkit-archive

cudnn：https://developer.nvidia.com/rdp/cudnn-archive

6. Install TensorFlow-GPU:

Install the latest TensorFlow-GPU version (replace 2.10.0 with the latest version):

pip install --upgrade tensorflow-gpu==2.10.0

python版本要求，不能过高也不能过低：https://tensorflow.google.cn/install/source?hl=zh-cn#tested_build_configurations

Verification:

1. Open Python: via cmd/ powershell.
   python
2. Import TensorFlow:
   import tensorflow as tf
3. Check GPU availability:
   print(tf.test.is_gpu_available())