https://raw.githubusercontent.com/MicrosoftLearning/mslearn-databricks/refs/heads/main/Instructions/Exercises/AI-04-Fine-tune-models.md - Azure behavior has changed

[SkillableKarl]

In relation to closed issue (#114) we would like to rewrite the ACP to match current behavior, however when running through this lab we get to Task 'Create a new notebook and ingest data' Step 6, and it asks us to populate '<catalog_name>', however I am uncertain what catalog to use. Recommended to re-run this lab in an allow all state and update instructions to clarify in a more detailed way exactly how to run this Task and subsequent steps.

[weslbo]

Hi @SkillableKarl , the instructions mention to use the default catalog. The name is typically the same as the name of azure databricks workspace.

[SkillableKarl]

That helps, although usage of the existing instructions could certainly be clarified with this very useful bit of information. I wonder if the CO can incorporate this? Also, it is likely a good idea for the CO to update the steps describing what exact model to use as it it appears gpt-4o - at least when following instructions verbatim, can lead to errors indicating the base model does not support fine-tuning: