In Machine Learning Systems Lab, we investigate programmability, scalability,  manageability, and performance of machine learning systems. We optimize the performance of existing machine learning systems such as Tensorflow, PyTorch, Megatron, and , as well as big data systems such as Hadoop. We investigate various trade-offs (such as performance and accuracy) in existing systems. We also work on security aspect of machine learning models and systems. See Research Page for the list of the research topics.

Research Topics include  (연구 주제):

• Large-Scale Machine Learning System             대규모 분산 머신러닝 시스템

• LLM Training/Inference Optimization                 LLM  학습/추론 최적화

• Deep Learning Model Security/Testing            딥러닝 모델 결함/편향성 탐지 및 개선

• Deep Learning Model Quantization                   딥러닝 모델 경량화/최적화

• Graph Processing Systems                                        그래프 분석 시스템

  자세한 내용은 Research Page 참고