What is missing is a definition of RAG. The whole page uses this acronym like it has to be clear what it is. It is not.
Not that people can't google it, but it is just friendlier if you answer the question a good chunk of people looking at the site will have. It also allows users who need what RAG provides but don't know what it is called to discover it easier.
It’s enough for the demo to use a variable named ‘kb’, which harkens to the business logic jargon. I don’t see benefit for expanding technical jargon once the business jargon has been uttered.
Chunking on hierarchy is a good and async built in and a cross encoder mode .... I like this project's Keep It Simple Stupid approach without skipping on functions even a basic graph triple. Using this to fill out a PoC mockup could be worth it vs dummy data and just drawing a cloud.
What is missing is a definition of RAG. The whole page uses this acronym like it has to be clear what it is. It is not.
Not that people can't google it, but it is just friendlier if you answer the question a good chunk of people looking at the site will have. It also allows users who need what RAG provides but don't know what it is called to discover it easier.
It’s enough for the demo to use a variable named ‘kb’, which harkens to the business logic jargon. I don’t see benefit for expanding technical jargon once the business jargon has been uttered.
Chunking on hierarchy is a good and async built in and a cross encoder mode .... I like this project's Keep It Simple Stupid approach without skipping on functions even a basic graph triple. Using this to fill out a PoC mockup could be worth it vs dummy data and just drawing a cloud.
Just tested it. Worked brilliantly well after fixing a minior issue.
My question: any plans on adding graph (rdf) support in near future?
Thanks..
Great documentation and dev-friendly feature set. Looks promising.
Nice, can you make it work with aws bedrock?