Multilayer Perceptron with Jobman

me@ryanorban.com (Ryan Orban) — Tue, 11 Feb 2014 00:00:00 +0000

Summary

This tutorial from deeplearning.net (the website of the LISA Lab at Université de Montréal, home of Yoshua Bengio’s group) shows how to train a multilayer perceptron (MLP) using Jobman, a job management system designed for running systematic hyperparameter search experiments. The combination reflects the research workflow at LISA Lab in 2013-2014: Theano for GPU-accelerated neural network computation, Jobman for orchestrating parallel experiments, and deeplearning.net as the public documentation and tutorial site.

Deep Learning Tutorials — DeepLearning.net

me@ryanorban.com (Ryan Orban) — Sun, 29 Dec 2013 00:00:00 +0000

Summary

The Deep Learning Tutorials site from LISA Lab at Université de Montréal was the definitive practical resource for learning deep learning in 2013. Written by researchers in Yoshua Bengio’s group, it provided code-first walkthroughs of core architectures using Theano — the Python framework the lab developed for symbolic math and GPU computation that preceded TensorFlow and PyTorch.

Theano on Ryan Orban

Multilayer Perceptron with Jobman

Summary

Deep Learning Tutorials — DeepLearning.net

Summary