https://github.com/enhuiz/langevin-dynamics-and-score-matching
A demo shows how to combine Langevin dynamics with score matching for generative models.
https://github.com/enhuiz/langevin-dynamics-and-score-matching
denoising-autoencoders langevin-dynamics score-matching
Last synced: about 2 months ago
JSON representation
A demo shows how to combine Langevin dynamics with score matching for generative models.
- Host: GitHub
- URL: https://github.com/enhuiz/langevin-dynamics-and-score-matching
- Owner: enhuiz
- License: mit
- Created: 2020-11-18T14:39:23.000Z (over 4 years ago)
- Default Branch: main
- Last Pushed: 2020-11-19T08:01:58.000Z (over 4 years ago)
- Last Synced: 2023-03-05T13:20:18.314Z (over 2 years ago)
- Topics: denoising-autoencoders, langevin-dynamics, score-matching
- Language: Python
- Homepage:
- Size: 23.9 MB
- Stars: 21
- Watchers: 3
- Forks: 4
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Langevin Dynamics and Score Matching
## How to?
### Setup
```
pip install -r requirements.txt
```
### Train
```
python3 main.py config/ring.yml train
```
This command produces the model (learnt scaled score function): `ring.pth`
### Test
```
python3 main.py config/ring.yml test
```
This command produces a list of frames at each stage of sampling under `ring/` and an animation `ring.gif` made from them.
## Demo (with carefully tuned hyperparameters)
- Left: sample from the real pixel distribution
- Right: sample using Langevin dynamics based on the score function learnt via denoising score matching.
### Ring
### Waddles
### Stewie
## Reference
- A good [lecture](https://youtu.be/3-KzIjoFJy4) on Langevin dynamics for sampling.
- A [recent paper](https://arxiv.org/pdf/1907.05600.pdf) combining Langevin dynamics with denosing score matching with a more complicated multi-level pertubation scheme.
- A [paper](http://www.iro.umontreal.ca/~vincentp/Publications/smdae_techreport.pdf) introducing the denoising score matching.