GitHub

MinerU-Diffusion: Rethinking Document OCR as Inverse Rendering via Diffusion Decoding

Our long-term goal is to build efficient and reliable 2.5B diffusion-based decoding for document OCR. MinerU-Diffusion reframes document OCR as an inverse rendering problem and replaces slow, ...
IEEE

DEval: Enhancing Large Language Models’ Decoding with Evaluation for Better Generation

Abstract: Existing Large Language Models (LLMs) generate texts through unidirectional autoregressive decoding methods to respond to various queries. These methods tend to consider token selection in a ...
marktechpost

Google AI Introduces STATIC: A Sparse Matrix Framework Delivering 948x Faster Constrained Decoding for LLM Based Generative Retrieval

In industrial recommendation systems, the shift toward Generative Retrieval (GR) is replacing traditional embedding-based nearest neighbor search with Large Language Models (LLMs). These models ...