修复两个训练失败根因: 1. conversations.json (665MB) 超过Node.js字符串限制导致ERR_STRING_TOO_LONG崩溃 2. 单文件崩溃导致整个训练管线中断,后续文件无法处理 - cos.js: 添加readBuffer/head/readPartial方法,cosRequest增加try/catch防止未捕获异常 - corpus-extractor-ops.js: 超大文件(>200MB)采用分块采样策略,ZIP文件提前检测避免无效读取 - cos-training-trigger.js: 增加文件大小预警日志,分类计数(成功/跳过/失败) Agent-Logs-Url: https://github.com/qinfendebingshuo/guanghulab/sessions/f10f1b02-ed4e-4ee8-ac27-0580c1cf8a7e Co-authored-by: qinfendebingshuo <207279273+qinfendebingshuo@users.noreply.github.com> |
||
|---|---|---|
| .. | ||
| age-os | ||
| app | ||
| nginx | ||
| proxy | ||
| scripts | ||
| setup | ||
| architecture.md | ||
| brain-server-profile.json | ||
| cn-server-profile.json | ||
| ecosystem.config.js | ||
| feimao-server-profile.json | ||
| migration-markers.json | ||
| migration-plan.md | ||
| persona-brain-architecture.md | ||
| zhuyuan-server-profile.json | ||