Deep Neural Networks are rapidly changing the world we live in today by providing intelligent data driven decisions. GPU’s have increasingly become the accelerator of choice for Deep Neural Networks. This module serves as an introduction to scientists who want to leverage the power of ROCm for accelerating DNNs.


Download the Presentation
➤  Running TensorFlow for MNIST Example
MNIST is a popular dataset used in the deep learning community. It has a corpus of handwritten digits that are identified by a trained deep neural network. This module walks through an example of running TensorFlow for MNIST.
Watch Video
Watch Video
Download the Lab

 

➤  Running PyTorch & LSTM Example
Long-Term Short Memory networks are useful for capturing time series data like speech and text. This tutorial will use a PyTorch example of World Level Language Modelling and train a multilayer LSTM.
Watch Video
See an example
Download the Lab

 

➤  Multi-GPU Deep Learning
Training Deep Neural Networks on multiple-GPUs is a common practice to enable faster training times as well as train larger models that cannot fit on a single GPU.
Watch Video
See an example
Download the Lab

 

➤  Wrap-up: Deep Learning on ROCm
Watch an overview of how Deep Learning is implemented using ROCm and the benefits of using ROCm in Deep Learning applications.
Watch Video