← 研踪 PaperTrailarXiv:2611.18459 · v1
Callo: Serving Cache-Miss Requests with Speculative Decoding on Invalidated KV under Prefill-Decode Disaggregation
首次提交 2026-11-14 01:28
版本更新 2026-11-14 01:28
雷达观测 2026-11-16 11:41
摘要
Agent sessions may involve frequent context edits, which effectively invalidate cached KV state and force requests to wait for re-prefill before decoding. We present Callo, a disaggregated serving system that leverages invalidated KV for speculative drafts to reduce this latency, while verifying drafts against the exact KV after prefill before committing tokens. Callo makes effective use of decode resources by scheduling speculations during periods of underutilization and preempting speculations when drafts are likely to not be accepted. We evaluated Callo on a two-GPU testbed using SWE-agent sessions across three model families, observing all-request speedups of 1.05-1.19x over vLLM at moderate load. We demonstrate that these gains are also realized at scale through cluster-level simulations driven by real-world traces.
作者备注:24 pages
观测到的版本
相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。
v1 · 2026-11-14 01:28 · 最新观测版本
Callo: Serving Cache-Miss Requests with Speculative Decoding on Invalidated KV under Prefill-Decode Disaggregation
Agent sessions may involve frequent context edits, which effectively invalidate cached KV state and force requests to wait for re-prefill before decoding. We present Callo, a disaggregated serving system that leverages invalidated KV for speculative drafts to reduce this latency, while verifying drafts against the exact KV after prefill before committing tokens. Callo makes effective use of decode resources by scheduling speculations during periods of underutilization and preempting speculations when drafts are likely to not be accepted. We evaluated Callo on a two-GPU testbed using SWE-agent sessions across three model families, observing all-request speedups of 1.05-1.19x over vLLM at moderate load. We demonstrate that these gains are also realized at scale through cluster-level simulations driven by real-world traces.
查看此版本解读 →查看此版本原文 ↗