A cross-device federated learning system, with the convergence process and device status visualized on a web browser.

Implementation for our paper DeepCache: Principled Cache for Mobile Deep Vision.

A benchmark tool to help developers understand the cost of on-device training, involving many typical deep learning models.

A tool to accurately detect deep learning apps that embed deep learning libraries and extract the models from those apps.

Breaking the Memory Wall for Resource-Efficient On-Device Machine Learning.

Artifacts for our SIGCOMM'22 paper "Multi-Resource Interleaving for Deep Learning Training"