FutureXiang/edm2

Minimal multi-gpu implementation of EDM2: "Analyzing and Improving the Training Dynamics of Diffusion Models"

18
/ 100
Experimental

This project helps machine learning researchers efficiently train and evaluate diffusion models for generating images. You provide a dataset of images, and it trains different configurations of diffusion models. The output is a trained model capable of generating new images, along with metrics like FID scores to assess generation quality. This is for researchers experimenting with cutting-edge image generation techniques.

No commits in the last 6 months.

Use this if you are an AI researcher wanting to implement and compare advanced diffusion model training techniques on image datasets like CIFAR-100.

Not ideal if you need to train diffusion models on very large-scale datasets like ImageNet-1k, as it would be computationally prohibitive.

AI-research image-generation diffusion-models machine-learning-engineering
No License Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 7 / 25
Maturity 8 / 25
Community 3 / 25

How are scores calculated?

Stars

40

Forks

1

Language

Python

License

Last pushed

Mar 05, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/FutureXiang/edm2"

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