# 混元大模型Infra稳定性研发工程师（深圳/北京/上海/杭州）

External opening: you cannot apply through AI Native Match; apply on the official site.

Apply through the official channel. The full external receipt stays in your workspace. When signed in, three minimal event types explicitly recorded through the CLI are reported by default: selection, self-reported application and channel observation. Browsing or saving a role does not report an event; external applications do not appear under My applications.

- Company: 腾讯 (Stage unknown)
- Locations: 深圳
- Role: Engineering
- Apply on the official site: https://careers.tencent.com/jobdesc.html?postId=1958073221844717568

## Role description from source

1.负责混元infra相关链路稳定性治理、规范建设；
2.联动框架、算力、网络各模块完善关键metric采集；
3.系统性构建故障节点、慢节点检测平台化能力；
4.联合混元一站式建设统一的任务自动续训能力；
5.响应并解决日常混元大模型任务的故障问题。
职位要求
1.熟悉Megatron/PyTorch等框架的基本的训练流程；
2.掌握GPU/NPU等工作原理、常见操作命令；
3.熟悉RDMA网络相关硬件特性、熟悉all2all、allGather等集合通信原理；
4.了解docker容器、存储挂载等基础知识；
5.有大规模任务系统故障排查、分析解决经验者优先；
6.良好沟通、团队协作能力。
工作年限（官方顶部字段）
一年以上工作经验

---
opening_id: 588cd85a-304b-4b59-8e72-6492b8549393

`anm openings save --id 588cd85a-304b-4b59-8e72-6492b8549393`