39 citations · 98 across the 8 of their papers we have counts for
4 papers · 1 filter
Elixir: Train a Large Language Model on a Small GPU Cluster
Haichen Huang, Jiarui Fang, Hongxin Liu +2
In recent years, large language models have achieved great success due to their unprecedented size. However, training these models poses a challenge for most researchers as it requ…
TurboTransformers: An Efficient GPU Serving System For Transformer Models
Jiarui Fang, Yang Yu, Chengduo Zhao +1
The transformer is the most critical algorithm innovation of the Nature Language Processing (NLP) field in recent years. Unlike the Recurrent Neural Network (RNN) models, Transform…
swCaffe: a Parallel Framework for Accelerating Deep Learning Applications on Sunway TaihuLight
Jiarui Fang, Liandeng Li, Haohuan Fu +5
This paper reports our efforts on swCaffe, a highly efficient parallel framework for accelerating deep neural networks (DNNs) training on Sunway TaihuLight, the current fastest sup…
RedSync : Reducing Synchronization Traffic for Distributed Deep Learning
Jiarui Fang, Haohuan Fu, Guangwen Yang +1
Data parallelism has become a dominant method to scale Deep Neural Network (DNN) training across multiple nodes. Since synchronizing a large number of gradients of the local model…