AI Tools & Models HOTFeaturedAll
arXiv cs.AIScore60

RL$^{\mathrm{RT}}$: On-Device Reinforcement Learning Runtime

Summary

This on-device RL runtime puts a periodic actor and a learner on a single embedded GPU, balancing their workloads so they share the hardware smoothly.

Why this matters

Relevant to this topic

Open original ↗

Source: arXiv cs.AI · arxiv.org

Related items