This repo is the work for the project of course Deep Learning 2.
Implementation of RBM, DBN, DNN, VAE and experiment on Binary Alpha Digit and MNIST dataset.
Run RBM code with
python RBM.py --show_img
Optional arguments:
--digit: list of digits/letters that RBM structure will learn. Value of elements in range [0, 35]. Example: --digit 3 5 7. Default [3]
--iter: number of training iteration. Default 1000.
--nb_img: number of images to generate. Default 3.
--show_img: Mark to show images after generation.
--save_img: Mark to save generated images to /experiments/RBM
Example of running code with full optional arguments:
python RBM.py --digit 6 10 --iter 800 --nb_img 5 --show_img --save_img
Run DBN code with
python DBN.py --show_img
Optional arguments: Same as RBM
Example of running code with full optional arguments:
python DBN.py --digit 6 10 --iter 10 --nb_img 5 --show_img --save_img