4 papers
Lagom: Unleashing the Power of Communication and Computation Overlapping for Distributed LLM Training
Guanbin Xu, ZhenGuo Xu, Yuzhe Li +4
Overlapping communication with computation is crucial for distributed large-model training, yet optimizing it - especially when computation becomes the bottleneck-remains challengi…
A Hassle-free Algorithm for Private Learning in Practice: Don't Use Tree Aggregation, Use BLTs
H. Brendan McMahan, Zheng Xu, Yanxiang Zhang
The state-of-the-art for training on-device language models for mobile keyboard applications combines federated learning (FL) with differential privacy (DP) via the DP-Follow-the-R…
Prompt Public Large Language Models to Synthesize Data for Private On-device Applications
Shanshan Wu, Zheng Xu, Yanxiang Zhang +2
Pre-training on public data is an effective method to improve the performance for federated learning (FL) with differential privacy (DP). This paper investigates how large language…
Efficient Language Model Architectures for Differentially Private Federated Learning
Jae Hun Ro, Srinadh Bhojanapalli, Zheng Xu +2
Cross-device federated learning (FL) is a technique that trains a model on data distributed across typically millions of edge devices without data leaving the devices. SGD is the s…