segmind/distill-sd
Segmind Distilled diffusion
Implements knowledge distillation training for Stable Diffusion using a teacher-student approach, where smaller U-Net models (sd_small with 579M parameters, sd_tiny with 323M) learn to mimic a larger teacher model while maintaining generation quality. Training combines output-level and feature-level distillation losses across U-Net blocks, integrated with Hugging Face Diffusers for seamless compatibility with standard text-to-image pipelines and fine-tuning workflows.
619 stars. No commits in the last 6 months.
Stars
619
Forks
39
Language
Python
License
—
Category
Last pushed
Oct 18, 2023
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/segmind/distill-sd"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
PrunaAI/pruna
Pruna is a model optimization framework built for developers, enabling you to deliver faster,...
bytedance/LatentSync
Taming Stable Diffusion for Lip Sync!
haoheliu/AudioLDM-training-finetuning
AudioLDM training, finetuning, evaluation and inference.
Text-to-Audio/Make-An-Audio
PyTorch Implementation of Make-An-Audio (ICML'23) with a Text-to-Audio Generative Model
sayakpaul/diffusers-torchao
End-to-end recipes for optimizing diffusion models with torchao and diffusers (inference and FP8...