谷歌与MIT发布Coco智能体平台
谷歌和MIT搞了个Coco平台，让AI架构师能同时优化硬件和模型，比传统方法快一倍。
谷歌与MIT联合发布Coco平台，供TPU架构师进行硬件与模型协同设计。该平台使用数百GB的实时仿真数据，这些数据模型在预训练阶段未见过。Coco将每次仿真记录归一化到关系数据库中，确保每个报告数值都来自SQL查询。平台目标是缩短架构师设置仿真和获取洞察的时间。
Great report from Google and MIT.
Co-designing ML models and the accelerators is an exciting research direction.
This report introduces Coco, an agent platform TPU architects use for hardware and model co-design. Each decision depends on hundreds of gigabytes of fresh simulation sweeps that no model saw during pretraining.
Coco registers every sweep in a normalized relational database, so each number an agent reports comes from a SQL query. Agents compose typed tools on their own, run recurring analyses such as comparing systems at matched execution configurations, and read the UI's navigation state as context.
The near-term goal is halving the time architects spend setting up simulations and getting insights from them.
Paper: https://t.co/MUwt2nkMOw
Chat with Paper: https://t.co/NR7BgGLCF1
