Lukaszl/pl-newspaper-pages-ocr-dataset-100-v1

Document OCR using GLM-OCR This dataset contains OCR results from images in Lukaszl/pl-newspaper-pages-ocr-dataset-100 using GLM-OCR, a compact 0.9B OCR model achieving SOTA performance. Processing Details Source Dataset: Lukaszl/pl-newspaper-pages-ocr-dataset-100 Model: zai-org/GLM-OCR Task: text recognition Number of Samples: 100 Processing Time: 11.2 min Processing Date: 2026-03-31 22:16 UTC Configuration Image Column: image Output Column:… See the full description on the dataset page: https://huggingface.co/datasets/Lukaszl/pl-newspaper-pages-ocr-dataset-100-v1.