Figures in pdf, with the relevant data in csv format for the resul figures in the paper "RAMP: A Flat Nanosecond Optical Network and MPI Operations for Distributed Computing and Deep Learning". I
We use ABCI (https://abci.ai/) for testing the effect of shuffling techniques on convergence of distributed SGD training This repository contains the artifact, e.g. job scripts, models, datasets.
Profiled data of DLRM models partitioned using table, column and data parallelism, used to estimate iteration time of large scale DDL workloads on multiple networks. This is used to evaluate the RAMP
