微信秒剪-多模态强化学习训练框架开发工程师
Core
Design, develop, and optimize training frameworks combining video generation models with reinforcement learning.
Role type
Senior IC multi-modal reinforcement learning training framework engineer
Builds
Optimized training pipelines for video generation models
Domain
AIGC, video generation, reinforcement learning
Deliverable
production ML models
Required skills
distributed training framework development, reinforcement learning theory and practice, video generation model principles, large-scale system optimization, multi-node multi-GPU communication optimization
Preferred skills
none stated
Technologies
none explicitly listed
Responsibilities
Design and develop training frameworks integrating video generation and reinforcement learning; Optimize multi-modal RL training workflows including environment setup and reward functions; Analyze and optimize full training pipelines for data loading and memory management; Resolve system bottlenecks in large-scale distributed training
Seniority
Senior, hands-on IC
