BumbleBee: Secure Two-party Inference Framework for Large Transformers

ePrint Report: BumbleBee: Secure Two-party Inference Framework for Large Transformers

Wen-jie Lu, Zhicong Huang, Zhen Gu, Jingyu Li, Jian Liu, Kui Ren, Cheng Hong, Tao Wei, WenGuang Chen

Large transformer-based models have realized state- of-the-art performance on lots of real-world tasks such as natural language processing and computer vision. However, with the increasing sensitivity of the data and tasks they handle, privacy has become a major concern during model deployment. In this work, we focus on private inference in two-party …

art bumblebee chen computer computer vision eprint report framework language large natural natural language natural language processing party performance report state tao tao wei transformers wei world