I'm kinda curious, why is LlamaIndex so

funny enough, it used to be called gpt-index, but the name got changed like a week for Llama LLM came out

In general, open-source LLMs are not great for a lot of the more intersting applications (agents, routing, structured outputs)

for basic q&a / rag, yea sure, they work

thats my general take anyways

thres a decent amount of support in the library these days for open-source stuff. Probably the biggest glaring issue is the defaults on openai, which can be sometimes frustrating to change

Hoping to fix that in the coming weeks by replacing the service context with proper global settings

I see, thanks for the response! Hoping the newer open-source models are able to meet those expectations. If not all from the same model, then certainly different specialized models for a given task.

I can only hope! 🙏

If it were my company, I'd named it Ragger, or something emphasizing on RAG 😂