大模型训练调优实习生
Tasks
- Analyze training stability and convergence
- Apply heterogeneous acceleration
- Build multi modal model training pipeline
- Collaborate on model structure and system bottleneck analysis
- Configure multi GPU resource scheduling
- Design training quality evaluation and anomaly monitoring
- Implement mixed precision training
- Implement multimodal data loading coordination
- Implement parameter efficient fine tuning
- Improve resource utilization efficiency
- Maintain pretraining finetuning evaluation deployment workflows
- Optimize training compilation
- Optimize training pipeline with task scheduling
- Use DDP distributed training
Perks/Benefits
- N/A
Skills/Tech-stack
Adapter | CI/CD | DDP | DeepSpeed | FSDP | LoRA | Multi-GPU | PEFT | PyTorch | Transformer
Education
Bachelor of Engineering | Bachelor of Science | Master of Science | PhD
Regions
Countries
States
Related jobs
-
Entry-level Full Time上海、深圳 R8h ago
-
MiMo-大模型训练框架开发工程师 CNY 240K-480KC++ | CI/CD | DeepSpeed | Distributed Training | GPU Memory OptimizationEntry-level Full Time北京 R7d ago
-
Mid-level Full Time北京 R7d ago
-
具身智能算法工程师-模型 CNY 500K-500KDeep learning | Distributed Training | IQL | Inference Optimization | Isaac LabMid-level Full Time北京 R7d ago
-
AI ML Engineer CNY 280K-360KAWS | Azure | C++ | Cloud Computing | Computer VisionPerformance bonuses | Professional development opportunities | Remote workMid-level Full TimeShenzhen, Guangdong Province, China R1mo ago
-
AI工程师-Agent Memory & RAG 方向（成都） CNY 240K-480KBERT | Chroma | Cross-Encoder | Embedding Models | FaissSenior-level Full Time成都 R1mo ago
-
AI工程师-Agent Memory & RAG 方向（武汉） CNY 240K-480KAlgorithms | BERT | Chroma | Cross-Encoder | Data StructuresSenior-level Full Time武汉 R1mo ago
-
Mid-level Full Time上海、深圳 R1mo ago