Question on providing image input

General
1

I want to input an image from my local and retreive content from a pdf. Based on the retreived conent from pdf and image input , i want to generate a summary . I sit possible in weaviate?

2

hi @SOHAIL_MOHD !!

Welcome to our community! :hugs:

It is. Check this multi modal tutorial:

While it only covers ingesting media (audio, image, video), you can also ingest your pdfs as text, and query the uploaded image filtering your objects by that single pdf (or multiple sources, if you want).

You can use for example a property called source with the name of the file.

Here you can find more information on the module used (multi2vec-bind):

In our recipes repo you can find integrations for llm-frameworks such as langchain for loading, chunking and indexing content from your pdfs.

Let me know if this helps!

Thanks!