(9/12) Does compressing activations help model parallel training?
Authors: @bian_song
, @DachengLi177
, Hongyi Wang, @ericxing
, Shivaram Venkataraman

→ View original post on X — @cohere