Hello and thanks for making this subreddit an amazing place to learn new things.
Are there resources online that describe how to use LocalLLaMA for OCR? In the past I’ve used OCRMyPDF to good effect, it does a solid job of pre/post processing + tesseract. I’ve uploaded a few documents that combine typed content + handwritten text to ChatGPT and it does an incredible job (exceeding all expectations). Is there anything beyond donut that I’m missing online that explains how this done?
You must log in or register to comment.
You could try LLaVA or MiniGPT-4 but I am unsure of how well they would perform.
Thanks, I’ll definitely check these out.