Repository navigation
Introduce Feast NLP/LLM Add-On #4964
Description
Activity
- added a sub-issue
on Feb 10, 2025 - changed the title
[-]Introduce Feast LLM Add-On[/-][+]Introduce Feast NLP/LLM Add-On[/+]on Feb 28, 2025 @ntkathole I've updated this to link to more issues.
CC @HaoXuAI @shuchu @ejscribner
Reacted by Nikhil KatholeThis issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
@franciscojavierarceo I dont have permissions to edit this issue. Could you add #5426 to sub-tasks as we did discuss on the slack thread (long time ago, but forgot to do it) ?
Reacted by Francisco Javier Arceodone!
This was implemented.
Is your feature request related to a problem? Please describe.
With #4364, #4890, and #4714, Feast is aiming to support NLP/LLMs as a first class citizen.
Other NLP related frameworks offer default transformations (e.g., formatting data via docling, llamaparse) and formatting data for ease of use in LLMs.
This should also include adding an NLP specific Feature Server as suggested in #5090
Feast should support this natively through an add on. The benefit of this add-on would be to allow users to have default libraries that can be used in offline training and online serving.
Describe the solution you'd like
To be scoped.
Describe alternatives you've considered
N/A
Additional context
https://docs.llamaindex.ai/en/stable/llama_cloud/llama_parse/
https://help.openai.com/en/articles/8868588-retrieval-augmented-generation-rag-and-semantic-search-for-gpts