I am having Issues with installing pip dependancies within my devImage. I was wondering if instead of using devImage: ghcr.io/loft-sh/devspace-containers/python:3-alpine I could use something like devImage: ghcr.io/loft-sh/devspace-containers/python:3.8-slim??
Or how would I just use my normal Docker image that I have made?
Specific issues with numpy/pandas and streamlit packages.
The documentation on devspace isn't very helpful.
I believe I need to change the config in my devspace.yaml file. If anyone has a good tutorial to do this on with python that would be amazing
I have tried installing lots of things with apk which may have worked however using alpine seems to be increibly slow.
FYI. I am on MacOS
Yes, you can absolutely use whatever image you want -- the example/pre-canned devspace containers that are published are just a starting point for you.
There is a python specific example here that covers everything you should need.
TL:DR -- build your image with whatever you want in it, and set that image as the image in your deployment and/or dev sections of the devspace.yaml file. You can have devspace build the image for you as well by setting the appropriate
image
block with your docker file/context/and image name.An example devspace.yaml: