Together Compute将演示推理平台升级：金丝雀部署、A/B测试及4倍加速预热
英文摘要
On August 6, Together Compute will host a live walkthrough of major updates to its inference platform designed to simplify production deployment of open-weight models. The new features include safe rollout strategies (canary, blue-green, and automatic rollback), A/B and shadow testing on live traffic, and SLO-driven autoscaling. The platform also now supports deploying any model with approximately 4x faster warm starts. The session will be led by inference product manager Nikitha and developer relations lead Zain Has.
中文摘要
8月6日，Together Compute将举办一场直播演示，展示其推理平台的重大更新，这些更新旨在简化开源权重模型的生产部署。新功能包括安全发布策略（金丝雀、蓝绿部署及自动回滚）、基于实时流量的A/B和影子测试，以及SLO驱动的自动伸缩。该平台现在支持任何模型约4倍更快的预热启动。会议将由推理产品经理Nikitha和开发者关系负责人Zain Has主持。
关键要点
Together Compute will host a live walkthrough on Thursday, August 6, 10:00–11:00 AM PDT.
Together Compute将于8月6日星期四太平洋夏令时上午10:00-11:00举办直播演示。
The platform now offers canary and blue-green deployments, automatic rollback, A/B and shadow testing on live traffic, and SLO-driven autoscaling.
平台新增金丝雀和蓝绿部署、自动回滚、实时流量A/B和影子测试，以及SLO驱动的自动伸缩功能。
Deploying any model now benefits from approximately 4x faster warm starts.
部署任何模型现在都能获得约4倍更快的预热启动速度。