大模型算法实习生（J93941）
Core
Train and align large language models, optimize inference, and build data flywheels to enhance model performance in real-world scenarios.
Role type
Intern, large model algorithm engineer
Builds
Production ML models, inference optimization tools, and data collection/analysis systems for AI Agents
Domain
Artificial Intelligence, Large Language Models, Deep Learning
Deliverable
production ML models
Required skills
Python, C++, PyTorch, Megatron, DeepSpeed, HuggingFace, Transformers, data structures, algorithm design, model training, model evaluation, PE optimization
Preferred skills
ACM/O/NOI/Top Coder awards, top conference papers, SOTA model exploration, system optimization
Responsibilities
Iterate data and train/evaluate models, optimize inference for specific scenarios, develop tools for AI Agent data collection and analysis, explore frontier modeling strategies to achieve SOTA results
