pbevan1 / llama-3.1-8b-ocr-correction

LLaMA 3.1-8B, finetuned on a synthetic OCR dataset for superior OCR correction.

20 runs
Public