完整原始转录 · 音频文稿点击展开
codex 果然有多烧 token 的问题
手上项目一周能烧 1 万刀的 token,样本够大了,因此让 codex 做了些统计,发现其中大量 token 浪费在了 codex 最近版本新增的 js tool 默认的高频等待上
解决方案:在官方修复前,可在全局 AGENTS.md 追加以下几行(新会话或老会话 compact 后生效)
For long-running asynchronous work:
- Empty `write_stdin` polls MUST use `yield_time_ms >= 180000`;
prefer `300000` when intermediate output is not needed.
- `functions.wait` MUST use `yield_time_ms >= 180000`.
- `functions.exec` MUST set its outer `@exec yield_time_ms` at least
30000 ms longer than the longest nested tool wait, so the outer
code cell does not yield first.
- Do not apply the long wait to non-empty `write_stdin` calls that
send interactive input.
- These tools return early when the process or cell completes.
Do not wake the model merely to report that work is still running.
加入这些 prompt 后,我的项目里统计下来节约了 25% 的 token
欢迎使劲转发把事情闹大 让自己的 codex 分析本地 session 进行验证[嘻嘻R] 根据经验不闹大点官方对那么多 issue backlog 当然是看也不看
#codex[话题]# #gpt[话题]# #openai[话题]#
codex 果然有多烧 token 的问题 手上项目一周能烧 1 ... https://xhslink.cn/o/7MOMQaSnVJd 带上口令,来【小红书】看笔记全文~
resolved_url: https://www.xiaohongshu.com/explore/6a7842ef000000002402e761
xhs_capture_status: success