微信搜索-大模型算法工程师-搜索 Planner 方向
Core
Design and develop algorithms for search planners using large language models to enhance query understanding, rewriting, and intent completion.
Role type
Senior IC large language model algorithm engineer (search planner)
Builds
Query understanding and rewriting systems, SFT/RLHF/DPO training pipelines, and efficient inference chains for search.
Domain
Search, Large Language Models, NLP
Deliverable
production ML models
Required skills
Transformer architecture, SFT/RLHF/DPO, LoRA/QLoRA, distributed training, DeepSpeed, Python, PyTorch, quantization, KV Cache optimization, AI-Native data production
Preferred skills
AI-Native workflow, automated pipeline construction, end-to-end business metric optimization
Technologies
PyTorch, DeepSpeed, LoRA, QLoRA
Responsibilities
Design query understanding and rewriting systems; build training data pipelines; solve large model training alignment issues; design efficient inference linkages; establish and optimize evaluation systems.
Seniority
Senior, hands-on IC
