When ML Training Cuts Through Congestion: Just-in-Time Gradient Compression via Packet Trimming.

Xiaoqi Chen, Shay Vargaftik, Ran Ben Basat

Browse the full HOTNETS paper archive.