Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
PyTorch implementation for Score-Based Generative Modeling through Stochastic Differential Equations (ICLR 2021, Oral)
| Date | Stars |
|---|---|
| 2026-07-30 | 2124 |
| 2026-08-10 | 2127 |
| 2026-08-18 | 2128 |
| 2026-08-19 | 2129 |
| 2026-08-25 | 2130 |
| 2026-08-27 | 2130 |
| 2026-09-15 | 2129 |
| 2026-09-17 | 2130 |
| 2026-09-18 | 2130 |
| 2026-09-19 | 2130 |
| 2026-09-21 | 2129 |
| 2026-09-22 | 2131 |
| 2026-09-28 | 2130 |
| 2026-10-12 | 2131 |
| 2026-10-13 | 2131 |
| 2026-10-14 | 2132 |
| 2026-10-19 | 2131 |
| 2026-10-21 | 2131 |
| 2026-10-24 | 2132 |
| 2026-11-04 | 2132 |
Today
— stars today
This week
— stars this week
This month
+2 stars this month
Momentum
0.0
growth rate 0.00%/day
# Score-Based Generative Modeling through Stochastic Differential Equations
[](https://paperswithcode.com/sota/image-generation-on-cifar-10?p=score-based-generative-modeling-through-1)
This repo contains a PyTorch implementation for the paper [Score-Based Generative Modeling through Stochastic Differential Equations](https://openreview.net/forum?id=PxTIG12RRHS)
by [Yang Song](https://yang-song.github.io), [Jascha Sohl-Dickstein](http://www.sohldickstein.com/), [Diederik P. Kingma](http://dpkingma.com/), [Abhishek Kumar](http://users.umiacs.umd.edu/~abhishek/), [Stefano Ermon](https://cs.stanford.edu/~ermon/), and [Ben Poole](https://cs.stanford.edu/~poole/)
--------------------
We propose a unified framework that generalizes and improves previous work on score-based generative models through the lens of stochastic differential equations (SDEs). In particular, we can transform data to a simple noise distribution with a continuous-time stochastic process described by an SDE. This SDE can be reversed for sample generation if we know the score of the marginal distributions at each intermediate time step, which can be estimated with score matching. The basic idea is captured in the figure below:
Our work enables a better understanding of existing approaches, new sampling algorithms, exact likelihood computation, uniquely identifiable encoding, latent code manipulation, and brings new conditional generation abilities (including but not limited to class-conditional generation, inpainting and colorization) to the family of score-based generative models.
All combined, we achieved an FID of **2.20** and an Inception score of **9.89** for unconditional generation on CIFAR-10, as well as high-fidelity generation of **1024px** Celeba-HQ images (samples below). In addition, we obtained a likelihood value of **2.99** bits/dim on uniformly dequantized CIFAR-10 images.
## What does this code do?
Aside from the **NCSN++** and **DDPM++** models in our paper, this codebase also re-implements many previous score-based models in one place, including **NCSN** from [Generative Modeling by Estimating Gradients of the Data Distribution](https://arxiv.org/abs/1907.05600), **NCSNv2** from [Improved Techniques for Training Score-Based Generative Models](https://arxiv.org/abs/2006.09011), and **DDPM** from [Denoising Diffusion Probabilistic Models](https://arxiv.org/abs/2006.11239).
It supports training new models, evaluating the sample quality and likelihoods of existing models. We carefully designed the code to be modular and easily extensible to new SDEs, predictors, or correctors.
## **Integration with 🤗 Diffusers library**
Most models are now also available in 🧨 Diffusers and accesible via the [ScoreSdeVE pipeline](https://huggingface.co/docs/diffusers/api/pipelines/score_sde_ve).
Diffusers allows you to test score sde based models in PyTorch in just a couple lines of code.
You can install diffusers as follows:
```
pip install diffusers torch accelerate
```
And then try out the models with just a couple lines of code:
```python
from diffusers import DiffusionPipeline
model_id = "google/ncsnpp-ffhq-1024"
# load model and scheduler
sde_ve = DiffusionPipeline.from_pretrained(model_id)
# run pipeline in inference (sample random noise and denoise)
image = sde_ve().images[0]
# save image
image[0].save("sde_ve_generated_image.png")
```
More models can be found directly [on the Hub](https://huggingface.co/models?library=diffusers&pipeline_tag=unconditional-image-generation&sort=downloads&search=ncsnpp).
## JAX version
Please find a JAX implementation [here](https://github.com/yang-song/score_sde), which additionally supports class-conditional generation with a pre-trained classifier, and resuming an evalutiExcerpt of 15,335 characters
Read on GitHub
7
Patrick von Platen
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:8a57821becdf60b7, topic:diffusion-models, readme:image generation, readme:inpainting
matched fp:8a57821becdf60b7, topic:pytorch