问题描述
我正在使用pip3 install tensorflow==1.8.0
,但是它不支持GPU.
I am using pip3 install tensorflow==1.8.0
, but it doesn't have GPU support.
所以我正在使用pip3 install tensorflow-gpu==1.8.0
,但是它仍然会引发异常
So I am using pip3 install tensorflow-gpu==1.8.0
, but it still raises an exception
我应该使用colab
从源代码安装tensorflow
吗?
Should I use colab
to install tensorflow
from source?
在pip3 list
之后:
tensorboard 1.10.0
tensorflow 1.10.0
tensorflow-hub 0.1.1
推荐答案
您可以将Tensorflow降级到以前的版本,而无需在Google Colab上提供GPU支持.我跑了:
You can downgrade Tensorflow to a previous version without GPU support on Google Colab. I ran:
!pip install tensorflow==1.12.0
import tensorflow as tf
print(tf.__version__)
最初返回的
2.0.0-dev20190130
但是当我在几个小时后返回时,得到了我要求的版本:
but when I returned to it after a few hours, I got the version I requested:
1.12.0
尝试降级到具有GPU支持的版本:
Trying to downgrade to a version with GPU support:
!pip install tensorflow-gpu==1.12.0
需要重新启动运行时并失败,因为导入import tensorflow as tf
会返回:
requires restarting the runtime and fails, as importing import tensorflow as tf
returns:
ImportError: libcublas.so.9.0: cannot open shared object file: No such file or directory
更新
导入失败时,您始终可以使用以下命令将CUDA降级到9.0版
Update
When the import fails you can always downgrade CUDA to version 9.0 using following commands
!wget https://developer.nvidia.com/compute/cuda/9.0/Prod/local_installers/cuda-repo-ubuntu1604-9-0-local_9.0.176-1_amd64-deb
!dpkg -i cuda-repo-ubuntu1604-9-0-local_9.0.176-1_amd64-deb
!apt-key add /var/cuda-repo-9-0-local/7fa2af80.pub
!apt-get update
!apt-get install cuda=9.0.176-1
您可以通过运行以下命令检查CUDA的版本:
You can check the version of CUDA by running:
!nvcc --version
第二次更新
此代码现在似乎失败了,请参见
这篇关于如何在Colab中降级Tensorflow版本?的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持!