coreyker/pendant

Python ENvironment for Deep Artificial Neural neTworks

Pendant

    .______   
    |   _  \  
    |  |_)  | 
    |   ___/  
    |  |      
    | _|      
          
P E N D A N T:  Python ENvironment for Deep learning with Artificial Neural neTworks

---

Description

Pendant is a simple python package for training and inference using (deep) artificial neural networks. This package was created for research and pedagogical purposes with musical data in mind (1d time-series data), however, many other good packges exist [1-3].

Currently supported models include restricted Boltzmann machines (RBM), multi-layer perceptrons (MLP), and deep neural networks (DNN).

Installation

Download/clone the repository and add its location to your PYTHONPATH environment variable. See the test scripts in each model file for example use-cases.

References

[1] Python: http://deeplearning.net/software/pylearn2

[2] Lua: http://torch.ch

[3] Matlab: https://github.com/dustinstansbury/medal