pytorch-energy-based-model

This repository provides simple illustrative working examples for energy-based models (EBM) in PyTorch.

The aim of the repository is to provide educational resources, to validate each step with toy examples, and to build a platform for future experiment.

Quickstart

The main requirements are python>=3.6 and torch>=1.2.

pip install -r requirements.txt

Validate Langevin dynamics sampling

python run_langevin.py 8gaussians

Training an energy-based model

python run_ebm.py 8gaussians

Expected Results

Directories

• run_langevin.py : Run Langevin dynamics sampling of a toy distribution. Produces images of samples.
• run_ebm.py : Train an EBM for a samples from a toy distribution.
• langevin.py : Codes related to Langevin dynamics
• model.py : Codes related to neural networks
• data.py : Codes related to generating toy distributions

Further reading

• IGEMB
• LeCun
• secretely