产品50
OpenAI 用 Codex 技能排查性能问题
原标题:Before Codex, engineers manually analyzed metrics and logs. To take that a step further, we captured our performance team’s investigation...
在 Codex 之前,工程师需要手动分析指标和日志。为了更进一步,我们把性能团队的调查过程封装成了一个名为“perf-forensics”的共享技能。这个技能告诉 Codex 有哪些可用工具、何时使用它们,以及如何排查变慢问题,从而让 Codex 能够解决该问题。 接下来,我们把调查迁移到云端,让告警可以自动启动一个 Codex 会话。 这帮助我们在构建 ChatGPT 的过程中保持节奏。
中文译文尚未完成,先为你展示原文。
Before Codex, engineers manually analyzed metrics and logs. To take that a step further, we captured our performance team’s investigation process in a shared skill called “perf-forensics”. This skill tells Codex which tools are available, when to use them, and how to work through a slowdown so Codex can resolve the problem.
Next, we moved investigations to the cloud so an alert could start a Codex session automatically.
That helps us keep pace as we build ChatGPT.
同一件事的其他报道1 篇
OpenAI 用 Codex 智能体守护 ChatGPT 性能
OpenAI Developers · 原文
