Pre$^3$: Enabling Deterministic Pushdown Automata for Faster Structured LLM Generation

TL;DR

Pre$^3$ introduces a method to convert LR(1) grammars into deterministic pushdown automata, significantly improving structured LLM generation efficiency by reducing token processing time and increasing throughput.

Contribution

It presents a novel approach to transform LR(1) transition graphs into DPDAs, enabling faster and more efficient structured output generation in LLMs.

Findings

Reduced time per output token by up to 40%.

Increased throughput by up to 36%.

Seamless integration into standard inference frameworks.

Abstract

Extensive LLM applications demand efficient structured generations, particularly for LR(1) grammars, to produce outputs in specified formats (e.g., JSON). Existing methods primarily parse LR(1) grammars into a pushdown automaton (PDA), leading to runtime execution overhead for context-dependent token processing, especially inefficient under large inference batches. To address these issues, we propose Pre$^3$ that exploits deterministic pushdown automata (DPDA) to optimize the constrained LLM decoding efficiency. First, by precomputing prefix-conditioned edges during the preprocessing, Pre$^3$ enables ahead-of-time edge analysis and thus makes parallel transition processing possible. Second, by leveraging the prefix-conditioned edges, Pre$^3$ introduces a novel approach that transforms LR(1) transition graphs into DPDA, eliminating the need for runtime path exploration and achieving edge transitions with minimal overhead. Pre$^3$ can be seamlessly integrated into standard LLM inference frameworks, reducing time per output token (TPOT) by up to 40% and increasing throughput by up to 36% in our experiments. Our code is available at https://github.com/ModelTC/lightllm.