Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/riannevdberg/sylvester-flows
https://github.com/riannevdberg/sylvester-flows
deep-learning normalizing-flows pytorch vae
Last synced: 3 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/riannevdberg/sylvester-flows
- Owner: riannevdberg
- License: mit
- Created: 2018-08-10T22:23:55.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2019-07-05T11:47:53.000Z (over 5 years ago)
- Last Synced: 2024-07-04T02:16:27.894Z (4 months ago)
- Topics: deep-learning, normalizing-flows, pytorch, vae
- Language: Python
- Size: 4.72 MB
- Stars: 176
- Watchers: 7
- Forks: 32
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Sylvester normalizing flows for variational inference
Pytorch implementation of Sylvester normalizing flows, based on our paper:
[Sylvester normalizing flows for variational inference](https://arxiv.org/abs/1803.05649) (UAI 2018)
Rianne van den Berg*, Leonard Hasenclever*, Jakub Tomczak, Max Welling
*Equal contribution
## Requirements
The latest release of the code is compatible with:
* `pytorch 1.0.0`
* `python 3.7`
Thanks to Martin Engelcke for adapting the code to provide this compatibility.
Version v0.3.0_2.7 is compatible with:
* `pytorch 0.3.0` **WARNING**: More recent versions of pytorch have different default flags for the binary cross entropy loss module: nn.BCELoss(). You have to adapt the appropriate flags if you want to port this code to a later vers\
ion.
* `python 2.7`
## Data
The experiments can be run on the following datasets:
* static MNIST: dataset is in data folder;
* OMNIGLOT: the dataset can be downloaded from [link](https://github.com/yburda/iwae/blob/master/datasets/OMNIGLOT/chardata.mat);
* Caltech 101 Silhouettes: the dataset can be downloaded from [link](https://people.cs.umass.edu/~marlin/data/caltech101_silhouettes_28_split1.mat).
* Frey Faces: the dataset can be downloaded from [link](https://github.com/y0ast/Variational-Autoencoder/blob/master/freyfaces.pkl).
## Usage
Below, example commands are given for running experiments on static MNIST with different types of Sylvester normalizing flows, for 4 flows:
**Orthogonal Sylvester flows**
This example uses a bottleneck of size 8 (Q has 8 columns containing orthonormal vectors).
```bash
python main_experiment.py -d mnist -nf 4 --flow orthogonal --num_ortho_vecs 8
```
**Householder Sylvester flows**
This example uses 8 Householder reflections per orthogonal matrix Q.
```bash
python main_experiment.py -d mnist -nf 4 --flow householder --num_householder 8
```
**Triangular Sylvester flows**
```bash
python main_experiment.py -d mnist -nf 4 --flow triangular
```
To run an experiment with other types of normalizing flows or just with a factorized Gaussian posterior, see below.
**Factorized Gaussian posterior**
```bash
python main_experiment.py -d mnist --flow no_flow
```
**Planar flows**
```bash
python main_experiment.py -d mnist -nf 4 --flow planar
```
**Inverse Autoregressive flows**
This examples uses MADEs with 320 hidden units.
```bash
python main_experiment.py -d mnist -nf 4 --flow iaf --made_h_size 320
```
More information about additional argument options can be found by running ```python main_experiment.py -h```
## Cite
Please cite our paper if you use this code in your own work:
```
@inproceedings{vdberg2018sylvester,
title={Sylvester normalizing flows for variational inference},
author={van den Berg, Rianne and Hasenclever, Leonard and Tomczak, Jakub and Welling, Max},
booktitle={proceedings of the Conference on Uncertainty in Artificial Intelligence (UAI)},
year={2018}
}
```