tyrell / llm-ollama-llamaindex-bootstrap-ui Goto Github PK

first off , thanks for your sharing this full stack open source project , after running the backend using python ```unicorn main:app --reload``,some errors happen as following:

 from app.api.routers.chat import chat_router
  File "/Users/yangboz/git/llm-ollama-llamaindex-bootstrap-ui/backend/app/api/routers/chat.py", line 6, in <module>
    from app.utils.index import get_index
  File "/Users/yangboz/git/llm-ollama-llamaindex-bootstrap-ui/backend/app/utils/index.py", line 2, in <module>
    from llama_index import VectorStoreIndex, ServiceContext
ImportError: cannot import name 'VectorStoreIndex' from 'llama_index' (unknown location)

any idea ? may be llama_index related ?

I'm trying to embeded both images and caption with RAG. I want the final RAG app to be able to receive images as input and provide a more context-aware answer (beyond the native multimodel model) based on the images and text in the vector store. How can we implement the image input here in the app ?
Thank you