ChainerMN: Scalable Distributed Deep Learning Framework
This work addresses the need for faster and larger-scale deep learning training for researchers and practitioners, though it is incremental as it builds on existing distributed computing concepts.
The authors tackled the challenge of scaling deep learning by developing ChainerMN, a distributed framework that achieved 90% parallel efficiency when training ResNet-50 on ImageNet across 128 GPUs.
One of the keys for deep learning to have made a breakthrough in various fields was to utilize high computing powers centering around GPUs. Enabling the use of further computing abilities by distributed processing is essential not only to make the deep learning bigger and faster but also to tackle unsolved challenges. We present the design, implementation, and evaluation of ChainerMN, the distributed deep learning framework we have developed. We demonstrate that ChainerMN can scale the learning process of the ResNet-50 model to the ImageNet dataset up to 128 GPUs with the parallel efficiency of 90%.