lvyufeng/denoising-diffusion-mindspore

Implementation of Denoising Diffusion Probabilistic Model in MindSpore

46
/ 100
Emerging

Provides a modular U-Net architecture with configurable timesteps and loss functions (L1/L2), integrated with MindSpore's mixed precision training (AMP) and exponential moving average (EMA) for stable convergence. Includes a high-level `Trainer` class supporting gradient accumulation, DDIM acceleration for faster inference, and direct dataset folder integration—eliminating boilerplate training code.

No commits in the last 6 months. Available on PyPI.

Stale 6m
Maintenance 0 / 25
Adoption 11 / 25
Maturity 18 / 25
Community 17 / 25

How are scores calculated?

Stars

46

Forks

10

Language

Python

License

MIT

Last pushed

Dec 16, 2022

Monthly downloads

21

Commits (30d)

0

Dependencies

1

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/lvyufeng/denoising-diffusion-mindspore"

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