With the human activity being a step where some manual validations, annotations, feedback might be required

Documentations is not very clear

It would be good if new agent package is published

I was thinking such limitations will exist only for published

Essentially. Not about removing all WARNINGS, but removing this as it actually works right and the WARNING is wrong.

This worked well:

if project_name is None and Task.current_task() is not None: project_name = Task.current_task().get_project_name()

AgitatedDove14 - any pointers on how to run gpu tasks with k8s glue. How to control the queue and differentiate tasks that need cpu vs gpu in this context

that worked, thanks AgitatedDove14

you will have to update this in your local clearml.conf, or wherever you are doing the Task.init from.

And anyway once a model is published can’t update it right? Which means there will be atleast multiple published models entries of same model over time?

Beyond this have the UI running, have to start playing with it. Any suggestions for agents with k8s?

Thanks! Is there GPU support, not clear from the Readme AgitatedDove14

Do let me know when task scheduler docs are up 🙂

tasks.add_or_update_artifacts/v2.10 (Invalid task status: expected=created, status=completed

yeah i was trying in local and it worked as expected. But in local I was creating a Task first and then seeing if it’s able to get project name from it

AgitatedDove14 aws autoscaler is not k8s native right? That's sort of the loose point I am coming at.

Am I missing something?

The image to run is empty essentially

Currently we train from Sagemaker notebooks, push models to S3 and create containers for model serving

Yeah

This is my code, but it’s pretty standard

SuccessfulKoala55 AgitatedDove14 - this is the example I recreated in demo server - https://demoapp.trains.allegro.ai/projects/ce59885cb4df4cd1afa34aa5b593bb54/experiments/58e02b02d1714f58a590e2335a0f2ab7/execution?columns=selected&columns=type&columns=name&columns=tags&columns=status&columns=project.name&columns=users&columns=started&columns=last_update&columns=last_iteration&columns=parent.name&order=-last_update

One thing I am looking at is nbdev from fastai folks

Will try set_parent, it wasn't in the docs and wasn't sure

Thanks AgitatedDove14 - is the datasets feature pretty new?

Great. Is there a good view of the roadmap?

BTW when I started using s3, I was thinking I needed to specify ouput_uri for each task. Soon realized that you just need the prefix where you want to put it into, and clearml will take care of project etc being appended to the path. So for most usecases, a single output uri set in conf should work.

As of now my understanding is agent and worker are just interchangeable words

AgitatedDove14 - tried exit(0) from the notebook and it worked

Thoughts AgitatedDove14 SuccessfulKoala55 ? Some help would be appreciated.