Code developped IFT6266 Deep Learning class at the university of Montreal. It is a wrapper for theano (https://github.com/Theano) symbolic math library to simplify development of deep learning models. It is strongly geared for computer vision applications.
It is split in two libraries
- libutils contains utilities for finding files in folders, displaying results.
- liblearn contains the theano wrapper code to make creation of deep nets.
Use dispatch.py for creating and dispatching experiments.