You can deploy this model and start earning money today!

zelaki/
eq-vae-ema

Name: eq-vae-ema
Availability: InStock

Views

No views yet

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

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

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

Loading the Model
You can load the model from the Hugging Face Hub:
python
1from transformers import AutoencoderKL 2model = AutoencoderKL.from_pretrained("zelaki/eq-vae-ema")

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