[Feature Request] Uploading PDFS/Text Files/Images?

[iChristGit]

I love the search function and it makes the chat feel so much more accurate! I use it mainly as a direct ChatGPT replacment, using code models when needed or normal models for chat.

Can we have the option to upload images/pdfs/other files to the chat? the images could be integrated by clip/blip, and the PDF or text files could just be added to the context or summarized and then added?

It would be awesome to have! Thank you for all the work made into this project

[nsarrazin]

Hi, we're working on it, we'll let you know as soon as it's done! 🧑‍🍳

[zacps]

I'm not sure if this is really the right place to comment, but I just wanted to document a similar feature request/use case.

We have an existing RAG engine which is highly customised to our particular use case. Currently this is behind a streamlit frontend, but we'd like to investigate moving this to chat-ui.

Currently we could implement the OpenAI protocol and provide this RAG engine as a model, but we'd have no way of configuring additional switches (for example which data sources to enable) or nicely rendering things like citations.

I'm not sure what this would look like for chat-ui, or if it's out of scope, but I wanted to put the use case out there.

[antonkulaga]

I am extremely confused: uploading is a feature that everyone needs and wants, some people already sent PRs with it but nothing works one year after. Is there a specific problem with this feature?