OCRia-1: A new SOTA 4B document parser
llamaindex.ai
2 threads
Interesting, got an issue in my project for OCR-ing images. I'll definitely try this out since ollama is supported anyways
would be cool to see energy usage associated with training