3 ms·What about using llama3.2-vision to do the OCR bit and then deferring to ChatGPT to do the CSV part?by mosselman 2y agoWhat about using llama3.2-vision to do the OCR bit and then deferring to ChatGPT to do the CSV part?