MiMo-大模型训练框架开发工程师
Tasks
- Accelerate data loading and preprocessing
- Analyze training performance bottlenecks
- Build performance monitoring system
- Design training framework
- Develop and optimize model training framework
- Implement distributed training
- Improve communication and computation overlap
- Optimize DP TP PP EP parallelism strategies
- Optimize memory and GPU utilization
- Tune training performance on different hardware
Perks/Benefits
- N/A
Skills/Tech-stack
C++ | CI/CD | DeepSpeed | Distributed Training | High Performance | High-Performance Computing | Infiniband | Megatron-LM | NCCL | Nsight Systems | Performance Computing | PyTorch | Python | RoCEv2
Education
N/A
Related jobs
-
AI基础设施研发工程师（Sandbox / 容器化)-MiMo CNY 180K-420KContainerd | Distributed Systems | Docker | ELK | File SystemMid-level Full Time北京 R1d ago
-
Mid-level Full Time上海、深圳 R4d ago
-
模型部署与推理优化工程师 CNY 180K-360KC++ | Edge inference | Inference Performance | Inference Performance Optimization | Model DistillationMid-level Full Time北京 R10d ago
-
Entry-level Internship上海 R11d ago
-
Entry-level Full Time北京 R13d ago
-
Mid-level Full Time北京 R13d ago
-
Mid-level Full Time北京 R13d ago
-
Mid-level Full Time北京 R28d ago
-
AWS | Azure | JavaScript | NoSQL | Node.jsFast-paced environment | Remote workMid-level Full TimeHangzhou R1mo ago
-
AWS | Agile | Azure | Blockchain | CursorRemote workMid-level Full TimeShenzhen R1mo ago