ytu-ce-cosmos/turkce-kitap

🔥 TurkishLLaVA OCR Enhancement Dataset This dataset is a specialized books collection designed to improve the Turkish OCR (Optical Character Recognition) abilities of the Turkish-LLaVA-v0.1 model. It was created by collecting 100,000 books entirely from Turkish sources. The primary goal of this dataset is to enhance the model's ability to detect and interpret any text present in images. Dataset Usage in Finetuning This dataset played a crucial role in the… See the full description on the dataset page: https://huggingface.co/datasets/ytu-ce-cosmos/turkce-kitap.