My first Krea2 Merge Model. As a w4a8 Smaller File with good results though.
Typical krea2 parameters are working. Use int8 Loader.
Giving thanks to my friend Tiwaz who helped me building this file version.
GonzalesKrea_V1bf16_00001 into_-INT8_CONVROT
gonzaleskreaV1w4a8_gonzaleskreaV1.safetensors