Is there a guide on how to update the base Docker image to necessary Python libraries for my team?
We have extra libraries that we would like to include natively in the Docker image allowing us to not have to pip install them every time the image spins up. I'm hopeful that this will save on average 30s or more of compute time every time we spin up a session.
Thanks in advance!
Modifying Docker Image
Comments
1 comment
While you can certainly pip install something while in a workspace (e.g. Jupyter notebook), if you'd like the package to persist between sessions we recommend that you customize your compute.
Here are some good documentations on how to do that: https://support.dominodatalab.com/hc/en-us/articles/115000392643-Environment-management
Here's a quick video summarizing the process: https://support.dominodatalab.com/hc/en-us/articles/360017555072
Submitted by: chuck.head
Please sign in to leave a comment.