This codebase contains a stacked autoencoder model. The main entrypoint is
run.py. See the documentation at the top of run.py for instructions on how
to train a model.
Configs specifying models are in the configs directory.
If you have questions about this implementation, please contact: Nick Watters, nwatters@mit.edu