Webinar 2015 Deep Learning on SHARCNET: From CPU to GPU cluster

Revision as of 10:29, 2 February 2016 by imported>Edward
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigationJump to search

In this webinar, a brief introduction of Deep Learning will be given. We will talk about the computational needs for some deep learning algorithms and discuss why they are more suitable for GPUs. Demos and benchmarks of the most popular deep learning software and libraries (Theano, Caffe, cuDNN, etc) will be shown on SHARCNET machines. We will also discuss what we should know about GPU clusters (hardware and software for CUDA-aware MPI) if we plan to build a deep learning model with over 10 billion trainable parameters.