soran-ghaderi/torchebm

🍓 Build and train energy-based and diffusion models in PyTorch ⚡.

54
/ 100
Established

Provides composable training objectives (contrastive divergence, score matching, equilibrium matching) and multiple sampling strategies—MCMC, optimization, and learned continuous-time dynamics via ODE/SDE integration. Includes pre-built analytical potentials, neural network architectures for conditional generation, synthetic benchmarking datasets, and numerical integrators for Hamiltonian dynamics. Targets PyTorch with CUDA acceleration and mixed precision support for efficient energy landscape exploration.

Available on PyPI.

Maintenance 10 / 25
Adoption 9 / 25
Maturity 25 / 25
Community 10 / 25

How are scores calculated?

Stars

81

Forks

7

Language

Python

License

MIT

Last pushed

Feb 17, 2026

Commits (30d)

0

Dependencies

3

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/generative-ai/soran-ghaderi/torchebm"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.