This is a light implementation of Denoising Diffusion Probabilistic Models on a toy manifold learning dataset s_curve from scikit-learn.
This code has tested with pytorch 1.13.1
See demo.ipynb
Visualization on the reverse process
- epoch = 0
- epoch = 1000
- epoch = 2000
- epoch = 3000
- epoch = 4000
- epoch = 5000
