Hi @<1886947834087870464:profile|SuperficialSparrow80>

You can configure the agent to run using conda in the configuration file, in the agent section - None

Hi @<1669152726245707776:profile|ManiacalParrot65> ,

Yes, you can wrap the separate function with a decorator so the function will run as a separate step in the pipeline, and even can cache the step for multi runs.

You can also add the function without a decorator, as a step to the pipeline with PipelineController.add_function_step() .

You can read about it [here](https://clear.ml/docs/latest/docs/pip...

@<1827878658367295488:profile|AnxiousOwl1> , currently, RBAC is an feature of the enterprise-tier ClearML Server....

Hi @<1797800418953138176:profile|ScrawnyCrocodile51> ,

I can understand the confusion here, from the code you shared, the issue is with how you're accessing the artifact properties. When you call .get() on an Artifact object, you're actually invoking the get() method defined in the Artifact class, which returns the artifact's content (in your case, a pandas DataFrame), not a dictionary value.
Instead, you should access the URL property directly:

task.artifacts.get("my-artifac...

@<1691983266761936896:profile|AstonishingOx62> agreed, will push it forward

thanks, checking it

Hi @<1790552666904989696:profile|GlamorousDog61> ,

Can you update clearml ( pip install -U clearml ) to the latest version?
Are you using add_function_step to add steps to the pipeline controller? if so, did you specify the packages with the packages parameter?

Hi @<1523708920831414272:profile|SuperficialDolphin93> ,

I would use Kibana to investigate the documents.