Hello ClearML community!
I have a question regarding batch inference using a base Docker image in ClearML. I have a Docker image that is configured with the necessary environment and code for the task at hand. My goal is to generate and enqueue a task utilizing this base image. However, it appears challenging to accomplish this without configuring the SOURCE CODE section (such as a git repository). I would like to avoid inspecting INSTALLED PACKAGES, as it is unnecessary for my prepared Docker container and seems to impede the task's startup speed.
Is there a way to achieve my inference task solely by utilizing the base image?
I appreciate your help in advance!