I guess it is a better then SDXL V1 but still not where I want it I guess there will be a V3...
Another one trained with Midjourney pictures