Responsibilities
- Design and build scalable systems to support large-scale model training
- Create and manage distributed training workflows for large language and multimodal models
- Enhance training efficiency across GPU clusters, multiple nodes, and data center environments
- Develop tools for monitoring, logging, and diagnosing training processes
- Design and manage storage architectures for massive training datasets
- Automate the setup, scaling, and coordination of training infrastructure
- Work closely with research teams to integrate and refine training techniques
- Evaluate and enhance the performance, scalability, and cost efficiency of training platforms
- Diagnose and resolve performance bottlenecks in distributed training setups
