Year: 2024
Type: preprint
Abstract: Parallel decoding methods such as Jacobi decoding show promise for more efficient LLM inference as it breaks the sequential nature of the LLM decoding process and transforms it into parallelizable computation. However, in practice, it achieves little speedup compared to traditional autoregressive (AR) decoding, primarily because Jacobi decoding seldom accurately predicts more than one token in a single fixed-point iteration step. To address this, we develop a new approach aimed at realizing fast convergence from any state to the fixed point on a Jacobi trajectory. This is accomplished by refining the target LLM to consistently predict the fixed point given any state as input. Extensive experiments demonstrate the effectiveness of our method, showing 2.4$\times$ to 3.4$\times$ improvements in generation speed while preserving generation quality across both domain-specific and open-domain benchmarks.
Source: arXiv (Cornell University)
Authors: Kou, Siqi, Hu, Lanxiang, He, Zhezhi, Deng, Zhijie, Zhang, Hao
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Awards: G396593899, 基于贝叶斯深度学习的偏微分方程解算子学习
OpenAlex ID: W4392537434