Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Bump docker images to use python 3.8+ #3303

Closed
creatorrr opened this issue Oct 3, 2022 · 4 comments · Fixed by #3701
Closed

Bump docker images to use python 3.8+ #3303

creatorrr opened this issue Oct 3, 2022 · 4 comments · Fixed by #3701
Assignees
Labels
P1 High priority, add to the next sprint topic:docker

Comments

@creatorrr
Copy link

The haystack docker images currently come with python3.7. Unfortunately, library support for <=3.7 is starting to wane. Is it possible to bump the python version for the images to 3.8 or greater? Maybe publish separate versions?

@masci
Copy link
Contributor

masci commented Oct 3, 2022

Hi @creatorrr

with 1.9 we deprecated the old docker images (see release notes) and the new ones already ship Python 3.10 in their CPU flavour.

The GPU version uses a Pytorch base image that unfortunately is still on Python 3.7.

We didn't have the time to see if we can safely update the Pytorch version we depend on, or which version of Python the more recent Pytorch images ship, but if you don't need GPU support you might be already good to go with the new 1.9 image.

@creatorrr
Copy link
Author

Gotcha, thanks for the update @masci . Unfortunately, I do need the gpu variants. Can you please point me to any specific docker build instructions that I can use? I'll try to see if I can get it built on base images with a more recent python version.

@masci
Copy link
Contributor

masci commented Oct 3, 2022

Have a look at https://github.com/deepset-ai/haystack/tree/main/docker it contains the bakefile to build the new images, you can take inspiration from there and try locally

@creatorrr
Copy link
Author

@masci opened a pr (wip) that uses nvidia/cuda as base and installs python3.10 manually on top of it

@masci masci added the P1 High priority, add to the next sprint label Nov 23, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
P1 High priority, add to the next sprint topic:docker
Projects
None yet
3 participants