Glow: Generative Flow with Invertible 1x1 Convolutions in Tensorflow 2

My implementation of GLOW from the paper https://arxiv.org/pdf/1807.03039 in Tensorflow 2. GLOW is an interesting generative model as it uses invertible neural network to transform images to normal distribution and vice versa. Additionally, it is strongly based on RealNVP, so knowing it would be helpful to understand GLOW's contribution.

Why?
Requirements
Training
Sampling
Future improvements
CONTRIBUTING
LICENSE

Why?

Normalizing flows is an interesting field of generative model as the optimization is derived from exact prior distribution of the images, as opposed to Variational Autoencoder (approximated using Evidence Lower Bound) and Generative Adversarial Network (using Jensen-Shannon Divergence).

The author of the paper has implemented the original version in Tensorflow 1 (https://github.com/openai/glow). However, with the current default version of Tensorflow 2, the repository is no longer actual. This brings the need of Tensorflow 2 implementation. Furthermore, here is provided the bare minimum of the algorithm which is easily modifiable. Simplicity is always the goal here and contribution is always welcome!

Note that the implementation is not exactly the same as what proposed in the paper mainly to improve the algorithm. This small differences lie in the network architecture and training hyperparameters.

Requirements

pip3 install -r requirements.txt

Training

After every epoch, the network's weights will be stored in the checkpoints directory defined in common_definitions.py. There are also some sampling of the network (image generation mode) that are going to be stored in results directory. Additionally, TensorBoard is used to track z's mean and variance, as well as the negative log-likelihood. In optimal state, z should have zero mean and one variance. Additionally, the TensorBoard stores sampling with temperature of 0.7.

python3 main.py [-h] [--dataset [DATASET]] [--k_glow [K_GLOW]] [--l_glow [L_GLOW]]
       [--img_size [IMG_SIZE]] [--channel_size [CHANNEL_SIZE]]

optional arguments:
  -h, --help            show this help message and exit
  --dataset [DATASET]   The dataset to train on ("mnist", "cifar10", "cifar100")
  --k_glow [K_GLOW]     The amount of blocks per layer
  --l_glow [L_GLOW]     The amount of layers
  --img_size [IMG_SIZE] The width and height of the input images
  --channel_size [CHANNEL_SIZE]
                        The channel size of the input images

More parameters of the implementation can be found at common_definitions.py. The pretrained weight for Cifar-10 can be downloaded at https://github.com/samuelmat19/GLOW-tf2/releases/download/0.0.1/weights.h5

Sampling

Sample the network with temperature of default 1.0

python3 sample.py [-h] [--temp [TEMP]]

optional arguments:
  -h, --help     show this help message and exit
  --temp [TEMP]  The temperature of the sampling

Future improvements

Improve documentation
Analyze instability of the network's training that occurs (Matrix Invertible when backpropagating to update weights)

CONTRIBUTING

To contribute to the project, these steps can be followed. Anyone that contributes will surely be recognized and mentioned here!

Contributions to the project are made using the "Fork & Pull" model. The typical steps would be:

create an account on github
fork this repository
make a local clone
make changes on the local copy
commit changes git commit -m "my message"
push to your GitHub account: git push origin
create a Pull Request (PR) from your GitHub fork (go to your fork's webpage and click on "Pull Request." You can then add a message to describe your proposal.)

LICENSE

This open-source project is licensed under MIT License.

CourierKyn/GLOW-tf2