# 大模型训练框架工程师（训练/推理/算子）

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: 阶跃 (Startup)
- Locations: Unknown
- Role: Engineering
- Apply on the official site: https://app.mokahr.com/social-recruitment/step/94904#/job/a48c4b47-c219-4932-aea0-19d67db5590c

## Role description from source

建设RL训练与推理设施、分布式任务调度和权重同步/热更新，优化吞吐、延迟、GPU利用率，并支持Agent RL。
熟悉PyTorch、CUDA、NCCL、Linux及分布式训推；理解RLHF或PPO/DPO/GRPO/Agent RL；有性能分析与复杂问题定位能力；有AI coding经验。

---
opening_id: 6c3b6382-4367-4670-8e75-2298c6d9d069

`anm openings save --id 6c3b6382-4367-4670-8e75-2298c6d9d069`