Pixel-space text-to-image generation (no VAE), monocular depth estimation and 4x image restoration/upscaling with Iris-3B by Speridlabs
No check-ins yet.