如何在 arch linux 中降级到 cuda 10.0？

Question

我想在 arch linux 中将我的 cuda 10.1 降级到 cuda 10.0，因为 tensorflow 只需要 cuda 10.0。

我在 arch 的 CUDA 10.1 上安装了 tensorflow Linux 但我不知道 tensorflow-gpu 需要 CUDA 10.0 或更低版本。

我尝试将所需的 10.0 库与新的 10.1 库进行符号链接，但没有成功。

每当我尝试在 python 中导入 tensorflow 时，控制台错误显示 'libcublas.so.10.0' 未找到。

所以我发现我只能在 CUDA 10.0 上运行 tensorflow 所以现在我想降级我的 CUDA。但我不知道如何在 arch Linux.

中做到这一点

每次我运行 import tensorflow 它都会给出这个错误。

Traceback (most recent call last):
  File "/home/techievin/.local/lib/python3.7/site-packages/tensorflow/python/pywrap_tensorflow.py", line 58, in <module>
    from tensorflow.python.pywrap_tensorflow_internal import *
  File "/home/techievin/.local/lib/python3.7/site-packages/tensorflow/python/pywrap_tensorflow_internal.py", line 28, in <module>
    _pywrap_tensorflow_internal = swig_import_helper()
  File "/home/techievin/.local/lib/python3.7/site-packages/tensorflow/python/pywrap_tensorflow_internal.py", line 24, in swig_import_helper
    _mod = imp.load_module('_pywrap_tensorflow_internal', fp, pathname, description)
  File "/usr/lib/python3.7/imp.py", line 242, in load_module
    return load_dynamic(name, filename, file)
  File "/usr/lib/python3.7/imp.py", line 342, in load_dynamic
    return _load(spec)
ImportError: /usr/lib/libcublas.so.10.0: version `libcublas.so.10.0' not found (required by /home/techievin/.local/lib/python3.7/site-packages/tensorflow/python/_pywrap_tensorflow_internal.so)

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/techievin/.local/lib/python3.7/site-packages/tensorflow/__init__.py", line 24, in <module>
    from tensorflow.python import pywrap_tensorflow  # pylint: disable=unused-import
  File "/home/techievin/.local/lib/python3.7/site-packages/tensorflow/python/__init__.py", line 49, in <module>
    from tensorflow.python import pywrap_tensorflow
  File "/home/techievin/.local/lib/python3.7/site-packages/tensorflow/python/pywrap_tensorflow.py", line 74, in <module>
    raise ImportError(msg)
ImportError: Traceback (most recent call last):
  File "/home/techievin/.local/lib/python3.7/site-packages/tensorflow/python/pywrap_tensorflow.py", line 58, in <module>
    from tensorflow.python.pywrap_tensorflow_internal import *
  File "/home/techievin/.local/lib/python3.7/site-packages/tensorflow/python/pywrap_tensorflow_internal.py", line 28, in <module>
    _pywrap_tensorflow_internal = swig_import_helper()
  File "/home/techievin/.local/lib/python3.7/site-packages/tensorflow/python/pywrap_tensorflow_internal.py", line 24, in swig_import_helper
    _mod = imp.load_module('_pywrap_tensorflow_internal', fp, pathname, description)
  File "/usr/lib/python3.7/imp.py", line 242, in load_module
    return load_dynamic(name, filename, file)
  File "/usr/lib/python3.7/imp.py", line 342, in load_dynamic
    return _load(spec)
ImportError: /usr/lib/libcublas.so.10.0: version `libcublas.so.10.0' not found (required by /home/techievin/.local/lib/python3.7/site-packages/tensorflow/python/_pywrap_tensorflow_internal.so)


Failed to load the native TensorFlow runtime.

pacman -S cuda=="10.0" 不工作。

我到处搜索，但找不到 arch Linux 降级到 CUDA 10.0

的任何正确答案

Answer 1

我面临同样的问题，必须 re-apply 我使用旧 gtx670 时使用的方法：

请记住，特定的 Cuda 版本需要特定的 nvidia 驱动程序。

看看 compatibility matrix 并最终阻止驱动程序升级。

Answer 2

从这里下载 cuda 10.0 包：http://archlinux.arkena.net/archive/packages/c/cuda/cuda-10.0.130-2-x86_64.pkg.tar.xz
安装包sudo pacman -U cuda-10.0.130-2-x86_64.pkg.tar.xz

Answer 3

我已经使用 yay. With yay you can install older versions of cuda and cudnn. In my experience this is faster than downloading from archive.org.

解决了这个问题

在 Arch 上安装 yay：

pacman -S --needed git base-devel
git clone https://aur.archlinux.org/yay.git
cd yay
makepkg -si

搜索您需要的版本：

yay -Ss cuda
yay -Ss cudnn

在撰写此答案时，我需要从 cuda 11.2 降级到 11.0。所以我安装了特定版本：

pacman -R cuda cudnn
yay -S cuda-11.0 cudnn8-cuda11.0

对我来说，降级后 Tensorflow 运行顺畅。

如何在 arch linux 中降级到 cuda 10.0？

How to downgrade to cuda 10.0 in arch linux?

cuda

archlinux

deep-learning

tensorflow