This repo supports a talk on extracting structured data from documents with OCR + LLMs.
- deepseek-ocr/: How to run deepseek-ocr on CPU (e.g. m1 Macbook)
- docling-extract-demo/: How to use Docling Extract to extract data from a document
- gemini-example/: How to use a foundational model to extract data via prompt from a document
- tesseract-example/: How to use a traditional OCR solution and extract data from the result