VSCode 中未检测到用于张量流的 GPU

Question

我已经为 Ubuntu 22.04 安装了 tensorflow 和 GPU 支持所需的依赖项，如 https://www.tensorflow.org/install/pip 中所述。

我测试了上面教程中提到的检测到 GPU，它检测到我的 Nvidia GTX 1060。但是如果我尝试在 VSCode 中的 GPU 上运行 tensorflow，则检测不到 GPU。

我有一个带有以下代码的 python 脚本test-tf.py：

import tensorflow as tf
print(tf.config.list_physical_devices('GPU'))

当我运行它时，我得到以下信息：

[PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU')]

当我在 VSCode 的 .ipynb 文件中运行相同的代码时，我得到：

[]
2022-12-27 18:59:38.079947: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2022-12-27 18:59:38.080624: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory
2022-12-27 18:59:38.080767: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcublas.so.11'; dlerror: libcublas.so.11: cannot open shared object file: No such file or directory
2022-12-27 18:59:38.080843: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcublasLt.so.11'; dlerror: libcublasLt.so.11: cannot open shared object file: No such file or directory
2022-12-27 18:59:38.080894: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcufft.so.10'; dlerror: libcufft.so.10: cannot open shared object file: No such file or directory
2022-12-27 18:59:38.080940: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcurand.so.10'; dlerror: libcurand.so.10: cannot open shared object file: No such file or directory
2022-12-27 18:59:38.080986: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcusolver.so.11'; dlerror: libcusolver.so.11: cannot open shared object file: No such file or directory
2022-12-27 18:59:38.081032: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcusparse.so.11'; dlerror: libcusparse.so.11: cannot open shared object file: No such file or directory
2022-12-27 18:59:38.081076: W tensorflow/compiler/xla/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
2022-12-27 18:59:38.081087: W tensorflow/core/common_runtime/gpu/gpu_device.cc:1934] 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...

如果我使用 Jupyter Notebook 运行相同的 .ipynb 文件，GPU 会按预期检测到。仅当从 VSCode 使用笔记本时才检测不到。我已确保使用适当的 conda 环境。

Answer 1

我遇到了同样的问题，但执行以下操作解决了我的问题

转到 vscode 中的终端选项卡-> 单击新终端

在左下角单击“+”号附近的下拉按钮，然后单击“设置默认配置文件”并选择“命令提示符”。

现在运行代码。

Answer 2

正如@TFer2 在问题的 cmets 中提到的，您需要从 conda 环境中打开 VSCode，并将 "terminal.integrated.inheritEnv": false 添加到用户 settings.json 中，tensorflow-gpu 才能使用 .ipynb 文件。