Hi LazyFox65
So the idea is that you add two lines of code to your codebase :
from clearml import Task task = Task.init(project_name='examples', task_name='change me')And you run it once, then it will create the experiment, environment arguments etc.
Now that you have it in the UI you can clone / change all the fields and send for execution.
That said you can also create an experiment from CLI (basically pointing to a repo and entry point)
You can read here:
https://github.com/allegroai/clearml/blob/master/docs/clearml-task.md

got it - so one repo with several tasks is useful, and not bad practice?

Correct 🙂

for example train.py & eval.py under the same repo

awesome. thx