zelaki/eq-vae-ema-ldm

Model

EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling

4 commits

3 linked in READMEs

updated Feb 25, 2025

See the code

README

EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling

Arxiv: https://arxiv.org/abs/2502.09509

EQ-VAE regularizes the latent space of pretrained autoencoders by enforcing equivariance under scaling and rotation transformations.

Model Description

This model is a regularized version of SD-VAE. We finetune it with EQ-VAE regularization for 44 epochs on Imagenet with EMA weights.

Model Usage

These weights are intended to be used with the EQ-VAE codebase or the CompVis Stable Diffusion codebase. If you are looking for the model to use with the 🧨 diffusers library, come here.

Metrics

Reconstruction performance of eq-vae-ema on Imagenet Validation Set.

Metric	Score
FID	0.552
PSNR	26.158
LPIPS	0.133
SSIM	0.725

Similar repos

zelaki/eq-vae-ldm

Model

EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling

zelaki/eq-vae-ema

Model

EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling

zelaki/eq-vae

Model

EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling

zelaki/eqvae

[ICML'25] EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling.

184

KBlueLeaf/EQ-SDXL-VAE

Model

EQ-SDXL-VAE: open sourced reproduction of EQ-VAE on SDXL-VAE

stabilityai/sd-vae-ft-ema

Model

Improved Autoencoders

136

stabilityai/sd-vae-ft-mse-original

Model

Improved Autoencoders

1,425

diffusers/sd-vae-ft-mse

Model

Improved Autoencoders

Linked in READMEs

CompVis/latent-diffusion

README badge

CompVis/stable-diffusion

A latent text-to-image diffusion model

73,491

README badge

zelaki/eqvae

[ICML'25] EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling.

184

README badge

zelaki/eq-vae-ema-ldm

Model

EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling

4 commits

3 linked in READMEs

updated Feb 25, 2025

See the code

README

EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling

Arxiv: https://arxiv.org/abs/2502.09509

EQ-VAE regularizes the latent space of pretrained autoencoders by enforcing equivariance under scaling and rotation transformations.

Model Description

This model is a regularized version of SD-VAE. We finetune it with EQ-VAE regularization for 44 epochs on Imagenet with EMA weights.

Model Usage

These weights are intended to be used with the EQ-VAE codebase or the CompVis Stable Diffusion codebase. If you are looking for the model to use with the 🧨 diffusers library, come here.