github/langchain
1
0
Fork 0
mirror of https://github.com/hwchase17/langchain.git synced 2025-09-12 21:11:43 +00:00
Code Issues Packages Projects Releases Wiki Activity

Update Gemini template README.md (#14993)

Browse Source
This commit is contained in:
Lance Martin
2023-12-20 20:07:20 -08:00
committed by GitHub
parent 8996d1a65d
commit 1db7450bc2
1 changed files with 7 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
templates/rag-gemini-multi-modal/README.md
View File

@@ -1,13 +1,15 @@
# rag-gemini-multi-modal
Multi-modal LLMs enable text-to-image retrieval and question-answering over images.
Multi-modal LLMs enable visual assistants that can perform question-answering about images.
You can ask questions in natural language about a collection of photos, retrieve relevant ones, and have a multi-modal LLM answer questions about the retrieved images.
This template create a visual assistant for slide decks, which often contain visuals such as graphs or figures.
This template performs text-to-image retrieval for question-answering about a slide deck, which often contains visual elements that are not captured in standard RAG.
It uses OpenCLIP embeddings to embed all of the slide images and stores them in Chroma.
This will use OpenCLIP embeddings and [Google Gemini](https://deepmind.google/technologies/gemini/#introduction) for answer synthesis.
Given a question, relevat slides are retrieved and passed to [Google Gemini](https://deepmind.google/technologies/gemini/#introduction) for answer synthesis.
![mm-mmembd](https://github.com/langchain-ai/langchain/assets/122662504/b9e69bef-d687-4ecf-a599-937e559d5184)
## Input
@@ -112,4 +114,4 @@ We can access the template from code with:
from langserve.client import RemoteRunnable
runnable = RemoteRunnable("http://localhost:8000/rag-gemini-multi-modal")
```
```