大模型研发工程师（J94734）
Core
Research and development of large model continuous pre-training and post-training, including data engineering, strategy, and effect evaluation.
Role type
Senior IC large model research engineer
Builds
Large language models via pre-training, supervised fine-tuning (SFT), and reinforcement learning (RL)
Domain
Artificial Intelligence / Large Language Models
Deliverable
production ML models
Required skills
Python, C++, CUDA, GPU system architecture, large model architecture design, large-scale data processing, data synthesis, model evaluation metrics
Preferred skills
Experience with pre-training, SFT, and RL projects
Technologies
Python, C++, CUDA, GPU systems
Responsibilities
Develop continuous pre-training and post-training pipelines for large models; troubleshoot technical issues during model training; engineer and strategize data collection, cleaning, and synthesis; evaluate model performance using various metrics
Seniority
Senior, hands-on IC
