大模型算法工程师（J85687）
Core
Develop and optimize large language models and visual large models for autonomous ride-hailing scenarios, focusing on SFT, prompt engineering, knowledge enhancement, and multimodal fusion.
Role type
Senior IC machine-learning engineer (large language models & visual models)
Builds
AI-native autonomous ride-hailing services
Domain
Autonomous driving / Transportation
Deliverable
production ML models
Required skills
large language model design, visual large model design, SFT, prompt engineering, knowledge enhancement, multimodal fusion, model selection, feature engineering, model training, model evaluation, model deployment, large-scale dataset preprocessing, data cleaning, feature extraction
Preferred skills
large model inference acceleration, high-performance inference deployment, distributed training (Deepspeed), fine-tuning frameworks (PEFT, LoRA, xtuner), online service deployment, resource scheduling (Kubernetes, Docker, GPU scheduling)
Technologies
Deepspeed, PEFT, LoRA, xtuner, Kubernetes, Docker
Responsibilities
Design, implement, and optimize model algorithms; preprocess and clean large-scale datasets for training; research cutting-edge large model technologies to support AI-native business reconstruction.
