Retrieval-Augmented Generation (RAG) with Spring AI, PGVector and Open AI

[@omkar.at.office]

Awesome explanation. Thanks!

[@Code.Wizzard]

Thanks a lot for the feedback

[@TomStephen-lz3jh]

Can you do a video with redis and using similarity search with filter expression?

[@Code.Wizzard]

Sure, will plan that. Thanks for the recommendation

[@liqwis9598]

Hello bro Can you do this with offline ai model using ollma

[@Code.Wizzard]

Sure, I'll plan a video for that. Thanks for the feedback! Meanwhile, if you want to try it out, it's really easy to swap OpenAI with Ollama. Simply change the dependency from spring-ai-openai-spring-boot-starter to spring-ai-ollama-spring-boot-starter. This video has instructions on how to set up Ollama locally: kzbin.info/www/bejne/d5-ymWN5ariXfLs

[@liqwis9598]

@@Code.Wizzard i have tried it out but not rag feature, so thought of asking you

[@Code.Wizzard]

That's great! Just switching the dependency in codebase of this video should do the trick then. However, while running Ollama locally, keep in mind that we usually use 7B or 13B models. This might lead to some reduction in accuracy compared to OpenAI's paid models, which use a much larger model, around 1.7 trillion parameters or so.

[@liqwis9598]

​@@Code.Wizzard let me try this 👍,, and bro please explain some more use. Cases with. Ai