I ask "what is TiDB" in the demo as suggested, and it takes 2 minutes to start responding in the midst of a multi-stage workflow with several steps each of graph retrieval, vector search, generation, and response combination.<p>Each of these is individually cool, but it strikes me as tragic that so much effort has been put into an intricate workflow and beautifully crafted UI only to culminate in a completely useless hello-world example, which after 5+ minutes of successive querying and response-building concludes with a network error.<p>I could use this to build exactly what I need...after stripping out 80% of the features to make it streamlined and responsive.<p>Why isn't that minimal version the default?
Is this wholly self-hostable? I'd be curious to run something like this on a home server, have some small model via ollama slowly chew through my documents / conversations / receipts / .... and provide a chat-like search engine over the whole mess.
Thanks a lot, this is the first time i saw a RAG using DSPy. I wanted to know about the expected cost. A few days ago fast graphrag compared their implementation with Microsoft:<p>> Using The Wizard of Oz, fast-graphrag costs $0.08 vs. graphrag $0.48 — a 6x costs saving that further improves with data size and number of insertions.
FYI the 'StackVM' link that pops up appears to show all inbound messages.<p><a href="https://stackvm-ui.vercel.app/tasks/3710e8d2-fb66-4274-9f78-42b1df80e9cb/master/aefe1205ff39cc0006787f0b82239010bc7424a4" rel="nofollow">https://stackvm-ui.vercel.app/tasks/3710e8d2-fb66-4274-9f78-...</a>
Oh, this looks pretty well made. Since it's using nextjs and shadcn/ui, I wonder if they also used v0 to generate components.<p>Has anyone any experience with TiDB? Haven't heard about it before this post
It looked neat but relies on a cloud db called 'TIDB', I checked its repo out and it looks like you can self host that as well but damn - it's a lot of containers. So yeah looks like self hosting is an option but likely a pain in the ass.