University College London
Browse

Profiled DDL RAMP

Download (1.27 GB)
dataset
posted on 2022-04-29, 09:56 authored by Alessandro OttinoAlessandro Ottino
<p>Collection of profiled models  used to estimate the disrtibuted training time for different Transformer Encoder models partiotioned using Megatron partitioning strategy, for different target losses</p>

History