This repository shares the latest deep learning example networks for training that focus on achieving the best performance and convergence from NVIDIA Volta Tensor Cores.

All of the examples shared here are also provided, together with our NVIDIA Deep Learning software stack, in a monthly updated docker container on the NGC registry (https://ngc.nvidia.com). These containers have the latest NVIDIA contributions sent upstream to the respective framework, as well as the latest NVIDIA Deep Learning software libraries, which have all been through a rigorous monthly quality assurance process to ensure that they provide the best possible performance. Additionaly, there are monthly release notes for each of the NVIDIA optimized containers: https://docs.nvidia.com/deeplearning/dgx/index.html#nvidia-optimized-frameworks-release-notes

The examples are organized first by framework and second by use case, such as computer vision, natural language processing, etc. We hope this structure better facilities quickly locating the options for networks that will best suite solving the challenge.

In each of the network READMEs, we indicate any known issues and encourage the community to help share any encountered issues.

In each of the network READMEs, we indicate the level of support that will be provided. The range is from ongoing updates and improvements to a point-in-time release for thought leadership.

We're posting these examples on GitHub to better support the community and facilitate feedback as well as contributions using GitHub Issues and Pull Requests. We welcome contributions!