Reproduce “Coda: Exploiting Admission Flexibility for Coding-Agent Serving” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Youhe Jiang, Fangcheng Fu, Binhang Yuan, Krishna Malladi, Ehsan K. Ardestani, Zhan Shu, Adnan Aziz, Yi Xu
Coda improves serving throughput for coding-agent requests by choosing which ready requests to prepare based on estimated state-preparation work and waiting time, while grouping attention for compatible contexts. Across tested single-worker and multi-worker configurations, it raises output-token throughput by an average 20.3% and SLO-compliant throughput by 70.5%. These results concern serving; experiments did not configure an SSD tier.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.