An Efficient Split Fine-tuning Framework for Edge and Cloud Collaborative Learning
This work addresses privacy and efficiency challenges in edge-cloud collaborative learning for NLP applications, representing an incremental improvement with novel techniques for communication reduction.
The paper tackles the problem of fine-tuning pre-trained models on edge devices without sharing local data with the cloud by proposing an efficient split fine-tuning framework, which reduces communication traffic by 96 times with minimal impact on accuracy across 9 NLP datasets.
To enable the pre-trained models to be fine-tuned with local data on edge devices without sharing data with the cloud, we design an efficient split fine-tuning (SFT) framework for edge and cloud collaborative learning. We propose three novel techniques in this framework. First, we propose a matrix decomposition-based method to compress the intermediate output of a neural network to reduce the communication volume between the edge device and the cloud server. Second, we eliminate particular links in the model without affecting the convergence performance in fine-tuning. Third, we implement our system atop PyTorch to allow users to easily extend their existing training scripts to enjoy the efficient edge and cloud collaborative learning. Experiments results on 9 NLP datasets show that our framework can reduce the communication traffic by 96 times with little impact on the model accuracy.