darioShar/DLPM

ICLR25 | Official code base for Heavy-Tailed Diffusion with Denoising Levy Probabilistic Models (DLPM)

26
/ 100
Experimental

This project offers an advanced generative model for creating new data similar to an existing dataset, particularly excelling with 'heavy-tailed' data or datasets where some categories are very rare. It takes your raw data, such as images or synthetic 2D points, and learns to produce high-quality, diverse new samples. Researchers and machine learning practitioners focused on data generation or augmentation for complex datasets would find this useful.

No commits in the last 6 months.

Use this if you need to generate realistic new data, especially from datasets with skewed distributions or where certain classes are underrepresented, and traditional methods fall short.

Not ideal if you primarily work with simple, evenly distributed datasets where standard generative models already perform sufficiently.

data-generation machine-learning-research image-synthesis dataset-augmentation generative-modeling
No License Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 7 / 25
Maturity 8 / 25
Community 11 / 25

How are scores calculated?

Stars

31

Forks

4

Language

Python

License

Last pushed

Feb 10, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/darioShar/DLPM"

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