Z-Image Base/Turbo Int8ConvRot quantizations for use with new Native ComfyUI INT8 support
Description
FAQ
Comments (2)
Thanks, would you mind sharing how you created this? Was it convert_to_quant by silveroxides or something else? And exactly what settings?
I am using the ComfyUI-INT8-Fast node to convert my models. I edited one line in the int8_save.py for native comfy use, as the original had no format naming, and edited the exclusion list in int8_unet_loader.py, best suited for turbo and base