Paper Info
Title
GPipe: Efficient Training of Giant Neural Networks using Pipeline Parallelism
Abstract
Scaling up deep neural network capacity has been known as an effective approach to improving model quality for several different machine learning tasks. In many cases, increasing model capacity beyond the memory limit of a single accelerator has required developing special algorithms or infrastructure. These solutions are often architecture-specific and do not transfer to other tasks. To address the need for efficient and task-independent model parallelism, we introduce GPipe, a pipeline parallelism library that allows scaling any network that can be expressed as a sequence of layers. By pipelining different sub-sequences of layers on separate accelerators, GPipe provides the flexibility of scaling a variety of different networks to gigantic sizes efficiently. Moreover, GPipe utilizes a novel batch-splitting pipelining algorithm, resulting in almost linear speedup when a model is partitioned across multiple accelerators. We demonstrate the advantages of GPipe by training large-scale neural networks on two different tasks with distinct network architectures: (i) Image Classification: We train a 557-million-parameter AmoebaNet model and attain a top-1 accuracy of 84.4% on ImageNet-2012, (ii) Multilingual Neural Machine Translation: We train a single 6-billion-parameter, 128-layer Transformer model on a corpus spanning over 100 languages and achieve better quality than all bilingual models.
Year
Field
DocType
2019
Computer science,Artificial intelligence,Artificial neural network,Machine learning
Conference
Citations 
PageRank 
References 
7
0.42
0
Authors
11
Name
Order
Citations
PageRank
Yanping Huang12109.80
Cheng, Youlong2191.28
Ankur Bapna3368.45
Orhan Firat428129.13
Dehao Chen5171.57
Xu Chen6305.73
HyoukJoong Lee741417.71
Jiquan Ngiam829719.56
Quoc V. Le98501366.59
Yonghui Wu10106572.78
Zhifeng Chen112747106.75