alamb commented on PR #14837:
URL: https://github.com/apache/datafusion/pull/14837#issuecomment-2872678278

   > Can we extend this sort of an approach to UDAFs? Having two entirely 
separate mechanisms would not be great.
   
   The data for an `async` user defined aggregate function must come from 
somewhere (presumably locally) and so aggregating using a remote service could 
be dicey 
   
   Maybe the usecaes is "send many rows of data to a remote LLM service" for 
example, 
   ```sql
   select user_id, call_llm_to_summarize(tweet_content) FROM tweets
   ```
   
   In this case, a lot of the user defined aggregate function API (like 
intermediate state, and GroupsAccumulators) doesn't make much sense. 


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: github-unsubscr...@datafusion.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscr...@datafusion.apache.org
For additional commands, e-mail: github-h...@datafusion.apache.org

Reply via email to