代码智能体算法工程师（J100658）
Core
Develop and optimize large language models (LLMs) for code agent applications, focusing on training, alignment, inference optimization, and data flywheel construction.
Role type
Senior IC machine-learning engineer (code agents)
Builds
Production code agents and optimized LLM inference systems
Domain
Artificial Intelligence / Large Language Models / Software Engineering
Deliverable
production ML models
Required skills
Python, C++, PyTorch, Megatron, DeepSpeed, HuggingFace, Transformers, algorithm design, system optimization, LLM training and alignment
Preferred skills
ACM/O/NOI/Top Coder awards, top conference publications, deep understanding of LLM underlying principles
Technologies
PyTorch, Megatron, DeepSpeed, HuggingFace, Transformers
Responsibilities
Iterate data and train/evaluate LLMs for code agent development; optimize inference and build data flywheels for real-world scenarios; improve code agent recognition accuracy; explore SOTA modeling and training strategies for code agents.