Python AutoEncoder from scratch using Numpy

Latent-space representation of the MNIST dataset using Variational Autoencoder

Usage

1. To install from source :

$ git clone git@github.com:lenoctambule/autoencoder.git
$ pip install -e autoencoder/

Or install from PyPI :

$ pip install easyvae

2. Optionally, run mnist_test.py to see it in action on the MNIST dataset.

$ cd examples
$ py mnist_test.py 

Training

Instatiate an ClassicalAutoencoder or VariationalAutoencoder object :

from easyvae.autoencoder import ClassicalAutoencoder, VariationalAutoencoder
from easyvae.activations import LeakyReLU

autoencoder = ClassicalAutoencoder(
    [768, 64, 16],
    [16, 64, 768],
    0.01,
    LeakyReLU()
)
# or
autoencoder = VariationalAutoencoder(
    [768, 64, 16],
    [16, 64, 768],
    0.01,
    LeakyReLU()
)

And then via the train_dataset method to train over a dataset :

autoencoder.train_dataset(data)

Or via the train method to input each data points iteratively :

autoencoder.train(v)

After training, you can save your model via the save method and load that model using load method :

autoencoder.save("mymodel.npy)
autoencoder.load("mymodel.npy")

Inference

Use your Autoencoder object with the encode, decode, forward methods like so :

example = ...
code = autoencoder.encode(example)
output = autoencoder.decode(code)
output, code = autoencoder.forward(example)