AIGC-模型训练加速工程师（北上杭深）
Core
Research and develop algorithms to optimize model training speed and efficiency using data/model parallelism and communication optimization.
Role type
Senior IC machine-learning engineer (training optimization)
Builds
Optimized training frameworks and strategies for core business AI models
Domain
AI infrastructure / Large Language Models / Computer Vision / NLP
Deliverable
production ML models
Required skills
distributed training frameworks, Python, PyTorch, FSDP, DeepSpeed, Megatron, model parallelism, data parallelism, communication optimization, performance bottleneck analysis, hardware utilization
Preferred skills
experience optimizing large-scale pre-trained models
Responsibilities
Analyze performance bottlenecks and implement optimization measures, maximize hardware resource utility, keep pace with industry technical progress
Seniority
Senior, hands-on IC