I guess that the container was destroyed with the data. is there any auto backup? mapping for local storage?

I restarted but now I have blank dashboard

do you know if I can restore the same container?

I updated the docker-compose and it fixed, thanks

looks great

I'll try

thanks
can I have it locally? cause with queue it's just pending

and the first Value error is still there..

thanks John, I'll set a demo soon

I thought maybe the experiment manager section can be integrated

no, I mean to launch clearml from kubeflow UI

not in use yet, but we want to take it as option in the future

organization solution

AgitatedDove14 As a user I excepted that task_name will be a str, so from my point of view, it should be explicit as same as in get_task.
one option is to validate task_name not contains any regex related chars in init phase, second option is to add explicit task_name option to the interface and not all-in-one option.

Thanks Martin.
I'll use this hack next time, but I think it might cause runtime issues when we use it naive way.
a different use for task_name will be more intuitive (especially when get_task works like this)

you can "clone" and "queue" tasks from the experiments UI but I'm not sure about time schedule from UI