deep-compression-autoencoder
Efficient vision foundation models for high-resolution generation and perception.