metadata
library_name: diffusers
pipeline_tag: unconditional-image-generation
tags:
- diffusion
- diffusion_model
Model Card for Model ID
This is my first test to train a diffusion model from scratch. It is composed of a simple UNet that denoises the noisy images. This is trained in a subset of the WHOI-PLankton dataset (only a part of the 2006 images). It was trained in a single Nvidia T4, ~6 minutes per epoch.