安装用于 CUDA 11.8 的 jaxlib

huangapple go评论180阅读模式
英文:

Installing jaxlib for cuda 11.8

问题

我正在尝试在我的Ubuntu 18上安装jax和jaxlib,使用Python 3.8,为了snerg(https://github.com/google-research/google-research/tree/master/snerg)。不幸的是,当我尝试使用以下命令安装Cuda 11.8版本的jax和jaxlib时:

pip install --upgrade jax jaxlib==0.1.69+cuda118 -f https://storage.googleapis.com/jax-releases/jax_releases.html

我遇到了以下错误:

ERROR: 忽略了需要不同Python版本的以下版本:0.4.14 Requires-Python >=3.9
ERROR: 找不到满足要求jaxlib==0.1.69+cuda118的版本(来自版本:0.1.32, 0.1.40, 0.1.41, ...)
ERROR: 未找到匹配的分发版本:jaxlib==0.1.69+cuda118

感激任何帮助。谢谢。

英文:

I'm trying to install jax and jaxlib on my Ubuntu 18 with python 3.8 for snerg (https://github.com/google-research/google-research/tree/master/snerg). Unfortunately when I try to install jax and jaxlib for Cuda 11.8 with the following command :<br/>

pip install --upgrade jax jaxlib==0.1.69+cuda118 -f https://storage.googleapis.com/jax-releases/jax_releases.html 

I get the following error:

ERROR: Ignored the following versions that require a different python version: 0.4.14 Requires-Python &gt;=3.9
ERROR: Could not find a version that satisfies the requirement jaxlib==0.1.69+cuda118 (from versions: 0.1.32, 0.1.40, 0.1.41, 0.1.42, 0.1.43, 0.1.44, 0.1.46, 0.1.50, 0.1.51, 0.1.52, 0.1.55, 0.1.56, 0.1.57, 0.1.58, 0.1.59, 0.1.60, 0.1.61, 0.1.62, 0.1.63, 0.1.64, 0.1.65, 0.1.66, 0.1.67, 0.1.68, 0.1.69, 0.1.70, 0.1.71, 0.1.72, 0.1.73, 0.1.74, 0.1.75, 0.1.76, 0.3.0, 0.3.2, 0.3.5, 0.3.7, 0.3.8, 0.3.10, 0.3.14, 0.3.15, 0.3.18, 0.3.20, 0.3.22, 0.3.24, 0.3.25, 0.4.0, 0.4.1, 0.4.2, 0.4.3, 0.4.4, 0.4.6, 0.4.7, 0.4.9, 0.4.10, 0.4.11, 0.4.12, 0.4.13)
ERROR: No matching distribution found for jaxlib==0.1.69+cuda118

Would appreciate any help. Thanks

答案1

得分: 2

jaxlib版本0.1.69相当老(它是在2021年7月发布的),而CUDA 11.8则在一年后的2022年9月发布。因此,我不会期望有针对jaxlib版本0.1.69的CUDA 11.8的预构建二进制文件。

如果可能的话,您最好安装一个更新的jaxlib版本,该版本包含针对CUDA 11.8的构建。当前的jaxlib+CUDA GPU安装说明可以在这里找到。

如果出于某种原因,您绝对需要这个非常老的jaxlib版本,那么您可能首先需要在您的系统上安装一个较旧的CUDA版本。jaxlib 0.1.69的CUDA安装说明可以在这里找到:看起来它是针对CUDA 10.1-10.2、11.0或11.1-11.3构建的。

英文:

jaxlib version 0.1.69 is quite old (it was released in July 2021) CUDA 11.8 was released over a year later, in September 2022. Thus I would not expect there to be pre-built binaries for jaxlib version 0.1.69 targeting CUDA 11.8.

If possible, your best bet would be to install a newer version of jaxlib, one which has builds targeting CUDA 11.8. The current jaxlib+CUDA GPU installation instructions can be found here.

If for some reason you absolutely need this very old jaxlib version, you'll probably first have to install an older CUDA version on your system. The CUDA jaxlib installation instructions from jaxlib 0.1.69 can be found here: it looks like it was built to target CUDA 10.1-10.2, 11.0, or 11.1-11.3.

huangapple
  • 本文由 发表于 2023年8月4日 04:04:49
  • 转载请务必保留本文链接:https://go.coder-hub.com/76831312.html
匿名

发表评论

匿名网友

:?: :razz: :sad: :evil: :!: :smile: :oops: :grin: :eek: :shock: :???: :cool: :lol: :mad: :twisted: :roll: :wink: :idea: :arrow: :neutral: :cry: :mrgreen:

确定