realtime-manglediffusion

ONNX (fp16) export of IDKiro/sdxs-512-dreamshaper (SDXS-512 DreamShaper, one-step, 0.32 B UNet, tiny VAE) for in-browser inference with onnxruntime-web / WebGPU. Used by the Stable Diffusion node's Realtime speed in MangleR.

Layout: tokenizer/, text_encoder/, unet/, vae_decoder/, vae_encoder/ (each model.onnx).

Weights are a conversion, not retrained. The OpenRAIL++ licence of the base model applies, including its use restrictions.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for szphynx/realtime-manglediffusion

Quantized
(5)
this model