![CUDA 工具包似乎已安装,但 tensorflow 没有看到 GPU](https://linux22.com/image/942929/CUDA%20%E5%B7%A5%E5%85%B7%E5%8C%85%E4%BC%BC%E4%B9%8E%E5%B7%B2%E5%AE%89%E8%A3%85%EF%BC%8C%E4%BD%86%20tensorflow%20%E6%B2%A1%E6%9C%89%E7%9C%8B%E5%88%B0%20GPU.png)
我有一台配备 RTX3060 的笔记本电脑,我在 ubuntu 20.04 上使用 tensorflow >= 2.6.0。我安装了 nvidia 驱动程序版本 495.44
我按照 tensorflow 安装了依赖项教程。我已经在 中安装了 cuda 11.5,/usr/local/cuda-11.5
并为 cuda 11.5 安装了 cudNN,并在 zshrc 中设置了路径。
当我在程序中导入 TensorFlow 时,没有任何警告/错误,但是当我尝试列出它们时,出现了以下情况:
tf.config.list_physical_devices('GPU')
2021-11-30 16:51:37.873669: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:939] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2021-11-30 16:51:37.895854: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudnn.so.8'; dlerror: libcudnn.so.8: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/cuda-11.5/include:/usr/local/cuda-11.5/lib64:
2021-11-30 16:51:37.895881: W tensorflow/core/common_runtime/gpu/gpu_device.cc:1850] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at https://www.tensorflow.org/install/gpu for how to download and setup the required libraries for your platform.
Skipping registering GPU devices...
[]
有人知道为什么 TensorFlow 仍然看不到我的 GPU 吗?