Databricks Community

@royinblr11 You're absolutely right to question the answer — the correct model for an application needing a foundation LLM with a large context window is: DBRXWhy DBRX is the Best Fit:It is a foundation model, designed for generation tasks.It support...

I still need help in this  

how can we turn on the playground option, i cannot find it

Hi @MaartenH,In Catalog and Schema names having space is not allowed by default while creation, even if notated with backsticks(``). Current best possibility is to rename the schema with only allowed values which you can find here for best practices ...

Hi,Please check if the cluster you are using to run the query as well upgraded to Unity Catalog. Also, follow the best practices outlined here for enablement: https://docs.databricks.com/aws/en/data-governance/unity-catalog/enable-workspacesBest Rega...

This has been dropped in favor of using a function defined internally within a notebook. If anyone has occasion to get a similar process set up - please let me know.Thanks

@drag7ter While I mentioned Workspace-level entitlements as an approach, I need to clarify something important: Databricks doesn't currently support creating fully customized entitlements where you can exclude specific permissions like "serving endpo...

Hi @jar Your scenario is quite common - managing costs while ensuring availability during business hours for Power BI Direct Query. Let me share some options based on my experience:Scheduled Cluster PoliciesThe most straightforward approach would be ...

Here are some suggestions:  1. Update coda.yaml. Replace the current config with this optimized version:  channels: - conda-forge dependencies: - python=3.10 # 3.12 may cause compatibility issues - pip - pip: - mlflow==2.21.3 - torch...

Hey greenPlatypus -  Free trials only offer serverless/SQL compute clusters (due to resource and cost controls). Standard or Premium tiers provide all-purpose compute clusters, and the Premium tier is recommended for full features and future proofing...

Vaibhavaher2025 -  I recommend trying the following:  1. Write logs from executors to persist storage insideprocess_partition. 2. Use mapPartitions instead offoreachPartition to return responses back to the driver as a Dataframe 3. Check executor log...

Can you please provide links? screenshot? more info? This answer is not specific enough.I'm taking the Data Analysis learning path, there are different demos I'd like to practice and there are no SP Lab environment links as mentioned in the videos.