MultiModalVectorStoreIndex.from_vector_store

[avenchen]

How to set the vector_store from MultiModal Qdrant store?

[dosubot[bot]]

🤖

Hello @avenchen! I'm Dosu, an AI bot designed to assist you with your LlamaIndex inquiries. While we wait for a human maintainer, I'm here to help sort out bugs, answer questions, and even guide you on becoming a contributor. Don't hesitate to ask questions - I'm here to help!

To set the vector_store from a MultiModal Qdrant store using MultiModalVectorStoreIndex.from_vector_store, you can follow the code snippet below:

from llama_index.indices.multi_modal.base import MultiModalVectorStoreIndex
from llama_index.vector_stores.types import VectorStore

# Assuming `vector_store` is your MultiModal Qdrant store
# and `service_context` is your ServiceContext instance

multi_modal_index = MultiModalVectorStoreIndex.from_vector_store(
    vector_store=vector_store,
    service_context=service_context,
    image_vector_store=None,  # Optional: You can provide an image_vector_store if you have one
    image_embed_model="clip"  # Optional: You can provide an image_embed_model if you have one
)

The from_vector_store class method is used to create an instance of MultiModalVectorStoreIndex from a VectorStore. It checks if the vector_store stores text, creates a StorageContext from the vector_store, and then returns an instance of MultiModalVectorStoreIndex. Please note that the vector_store should be a VectorStore that stores text. If it doesn't, a ValueError will be raised.

You can find more details about the MultiModalVectorStoreIndex class and the VectorStore protocol in the LlamaIndex codebase at the following links:

• MultiModalVectorStoreIndex
• VectorStore

Please let me know if you need further assistance.

Sources

• llama_index/indices/multi_modal/base.py
• llama_index/vector_stores/types.py

---

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.