Not-a-Bandit: Provably No-Regret Drafter Selection in Speculative Decoding for LLMs
This work addresses the challenge of efficiently selecting draft models in speculative decoding to accelerate LLM inference, which is incremental as it builds on existing speculative decoding methods.
The paper tackles the online draft model selection problem in speculative decoding for LLM inference, designing an algorithm that provably competes with the best draft model in hindsight per query and improves exponentially over bandit-based approaches as the number of draft models increases. Experiments show it substantially outperforms state-of-the-art methods like EAGLE3 and BanditSpec, especially in domains requiring long reasoning chains.
Speculative decoding is widely used in accelerating large language model (LLM) inference. In this work, we focus on the online draft model selection problem in speculative decoding. We design an algorithm that provably competes with the best draft model in hindsight for each query in terms of either the token acceptance probability or expected acceptance length. In particular, we show that we can accurately evaluate all draft models, instead of only the chosen model without incurring additional queries to the target model, which allows us to improve exponentially over the existing bandit-based approach as the number of draft models increases. Our approach is generically applicable with any speculative decoding methods (single draft, multi-drafts and draft-trees). Moreover, we design system-efficient versions of online learners and demonstrate that the overhead in computation and latency can be substantially reduced. We conduct extensive experiments on open-source LLMs and diverse datasets, demonstrating that our methods substantially outperform the state-of-the-art EAGLE3 and the BanditSpec baseline in a variety of domains where specialized domain-expert drafters are available, especially when long reasoning chains are required.