Upload files.

[solitude-alive]

Select Topic Area

Question

Body

I'm using a model such as GPT-4, and I'm wondering how I can upload a file, like a PDF, and ask the model to summarize or extract information from it.
This is straightforward to do with the ChatGPT client, but how can I achieve the same functionality with GitHub Models?

[pathanin-kht]

Hello @solitude-alive, I think to use GPT-4 with a PDF, you need to extract the text using libraries like PyPDF2 and send it to the model via an API with instructions like "Summarize this text." Unlike ChatGPT’s file upload feature, this requires coding the file processing and API setup.

[imDarshanGK]

Hi @solitude-alive,
GitHub Models doesn’t support direct file uploads. You can extract text from a PDF using a tool (like PyPDF2), then send that text to the model for summarization or info extraction via the API.

[solitude-alive]

Hi @pathanin-kht and @imDarshanGK ,
Thank you so much for your answer! I'll give it a try. By the way, if the PDF contains a lot of images and charts, do you have any good methods for handling those?

[pathanin-kht]

I haven’t tried it myself, but I think this could work. If your PDF has a lot of images or charts, you can use libraries like PyMuPDF or pdf2image to extract them. For charts, tools like OpenCV might help if they’re images. If they’re in vector format, PyMuPDF should handle it.

[danielg25]

Can you provide a detailed guide on how to use PyMuPDF or pdf2image to extract images from a PDF? Additionally, what is the most effective approach to handle charts in PDFs if they are images or in vector format?

[pathanin-kht]

Hi @kaunuto @danielg25,

I'm studying this, but I think to extract images from a PDF, you can use PyMuPDF or pdf2image. For charts, if they're images, Tesseract or OpenCV might help. If they're in vector format, PyMuPDF can export them as SVG files, and you can convert them with tools like cairosvg.

I'm not entirely sure, but I think this might help.

[imDarshanGK]

Hi,

1. Text Extraction: Use PyPDF2 to get the text from a PDF, then send it to GPT-4 using an API for summarization or information extraction.
2. Handling Images and Charts in PDFs:

• Use PyMuPDF or pdf2image to extract images.
• For charts as images, try Tesseract or OpenCV.
• If charts are in vector format, use PyMuPDF to export them as SVG and convert with cairosvg.