Feature request: RAG for search and chat

taichimaeda / markpilot

AI-powered inline completions and chat view for Obsidian

MIT License

6 stars 1 forks source link

Feature request: RAG for search and chat #5

Open taichimaeda opened 2 months ago

taichimaeda commented 2 months ago

Ollama for local embedding
RAG algorithm similar to LlamaIndex.TS
IndexedDB for vector store (currently not supported by LlamaIndex.TS)

Links:

https://ollama.com/blog/embedding-models https://ts.llamaindex.ai/

ythefly commented 2 months ago

hi I know my ollama connection is active and serving. When i connect to http://127.0.0.1:11434 in the web browser it said ollama is active and serving. But when i put that into the obsidian plugin it always said time out or not connected. What is wrong and how do i fix it?