大模型算法实习生（J106975）
Core
Building intelligent data infrastructure and embodied intelligence systems using large-scale multimodal models.
Role type
Intern, Multimodal AI Engineer (LLM/VLM/Embodied AI)
Builds
Multimodal operators, workflows, and fine-tuned LLM/VLM models for project deployment
Domain
Artificial Intelligence, Multimodal Learning, Embodied AI
Deliverable
production ML models
Required skills
Python, PyTorch, PaddlePaddle, LLM/VLM fine-tuning, CV model training, k8s, 3D reconstruction, SLAM
Preferred skills
Open-source model reproduction, workflow optimization
Technologies
PyTorch, PaddlePaddle, k8s, Python
Responsibilities
Develop high-quality multimodal operators and workflows; Fine-tune LLMs and VLMs; Build automated evaluation systems; Optimize operator efficiency for project deployment
