RL$^{\mathrm{RT}}$:端侧强化学习运行时,将截止期错过率降至 1% 以下
RL$^{\mathrm{RT}}$: On-Device Reinforcement Learning Runtime
AI 导读
arXiv 论文提出 RL$^{\mathrm{RT}}$,一个端侧强化学习运行时,把执行器与学习器的耦合变成可调参数:用 refresh bound 限制执行器释放,并用带 slack-aware admission 的 GPU permit 优先执行器、将空闲时间用于训练。
正文 · 原文
Abstract:On-device reinforcement learning runs a periodic actor and a learner on one embedded GPU shared with other workloads. Because the learner produces the actor's weights, existing designs stall the actor or let its policy drift. We present RL$^{\mathrm{RT}}$, an on-device runtime that makes this coupling a tunable parameter: a refresh bound limits actor releases between refreshes, and a GPU permit with slack-aware admission prioritises the actor and fills spare time with training atop any base scheduler. Under the real refresh cost on two Jetson boards, RL$^{\mathrm{RT}}$ keeps a 16.8 M-parameter policy's deadline misses below 1%, against 85 to 100% without it.
| Comments: | 8 pages, 11 figures |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.18035 [cs.LG] |
| (or arXiv:2611.18035v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2611.18035 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Zexin Li [view email]
[v1]
Fri, 13 Nov 2026 11:05:04 UTC (1,034 KB)
来源:arXiv cs.AI · arxiv.org