Add CC-scale MinerU-HTML layout-clustering + propagation pipeline (91% fewer LLM calls, F1=0.91)#2075

VibhuJawa wants to merge 118 commits into

NVIDIA-NeMo:mainNVIDIA-NeMo/Curator:mainfrom

VibhuJawa:feat/mineru-html-layout-clustering-pipelineVibhuJawa/NeMo-Curator:feat/mineru-html-layout-clustering-pipelineCopy head branch name to clipboard