Predictive Speculative KV Replication for Bursty LLM Inference

Hacker News Top Papers

Summary

A paper introducing predictive speculative KV replication to handle bursty LLM inference workloads, with code available on GitHub.

<a href="https:&#x2F;&#x2F;github.com&#x2F;jwlaboratory&#x2F;bite-the-bullet" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;jwlaboratory&#x2F;bite-the-bullet</a>
Original Article

Similar Articles

A Sparse Glimpse of the Whole: Train-Free Self-Speculative Decoding

arXiv cs.CL

This paper introduces SparseSpec-L, a training-free self-speculative decoding framework for long-context LLM inference that uses a dynamically sparsified and recallable KV cache along with an entropy-based controller for speculation length, achieving up to 2.79x speedup over autoregressive decoding.