豆包大模型算法工程师（火山方舟）-RL Infra
Core
Improving RL training systems and optimizing performance/stability for large model post-training tasks (Reasoning, Agent, VLM).
Role type
Senior IC large model RL infrastructure engineer
Builds
RL training systems and post-training pipelines for large models
Domain
Large Language Models / Reinforcement Learning / Distributed Systems
Deliverable
production ML models
Required skills
Reinforcement Learning algorithms, Distributed training frameworks, Python, C/C++/Rust/Golang/Java, Data structures and algorithms, High-performance inference engines
Preferred skills
OpenRLHF, VeRL, PyTorch FSDP2, DeviceMesh, DTensor, vLLM, SGLang, Continuous Batching, PagedAttention, Prefix Caching, SRFT, OnPolicy Distillation
Responsibilities
Extend RL Trainer functionality based on Ray, explore Rollout/sampling strategies, integrate Reward systems, manage trajectories in complex Agent Loop tasks, optimize SFT/RL training performance and stability, explore frontier algorithms like Off-Policy RL and DPO/PPO/GRPO
Seniority
Senior, hands-on IC
