Kymata Labs/The Living IndexesBuilt by tekvisions ↗
The Document Index / OCR Engines / #54
opendatalab

opendatalab/MinerU-Diffusion

by opendatalab · OCR Engines · updated 1mo ago

A diffusion-based framework for document OCR that replaces autoregressive decoding with block-level parallel diffusion decoding.

57
momentum
597
stars
38
forks
#54
rank
ai4sciencediffusiondlmdocument-analysisextract-datalayout-analysislladaocrparserpdfpdf-converterpdf-extractor-llm
View on GitHub →