广告架构高级工程师-Applied AI
Core
Design and support large-scale distributed training and inference for 100B+ parameter models with 1M sequence lengths, optimizing ML efficiency and supporting specialized paradigms.
Role type
Senior Applied AI Infrastructure Engineer
Builds
Distributed training and inference systems for massive language models
Domain
AI Infrastructure / Large Language Models
Deliverable
production ML models
Required skills
C/C++, Python, Linux, Distributed Training (DDP, FSDP, TP, SP, PP), Inference Frameworks (vLLM, TRT), Model Optimization (Quantization, Pruning, Distillation, NAS), Compiler Optimization
Preferred skills
GPU Programming, Reinforcement Learning, Federated Learning, Graph Learning, Optimization Algorithms, Machine Learning Frameworks (TensorFlow, PyTorch, Jax)
Technologies
vLLM, TensorRT, PyTorch, TensorFlow, Jax, Linux
Responsibilities
Implement distributed training and inference support for 100B+ parameter models; Collaborate with algorithm teams for joint optimization; Optimize ML efficiency via quantization, pruning, and compression; Support specialized ML paradigms like reinforcement and federated learning.
Seniority
Senior, hands-on IC