Lykon/AAM_XL_AnimeMix is a Stable Diffusion model that has been fine-tuned on
stabilityai/stable-diffusion-xl-base-1.0.
For more general information on how to run text-to-image models with 🧨 Diffusers, see the docs.
1from diffusers import AutoPipelineForText2Image, DEISMultistepScheduler
2import torch
3
4pipe = AutoPipelineForText2Image.from_pretrained('Lykon/AAM_XL_AnimeMix', torch_dtype=torch.float16, variant="fp16")
5pipe.scheduler = DEISMultistepScheduler.from_config(pipe.scheduler.config)
6pipe = pipe.to("cuda")
7
8prompt = "anime girl, night, blue light behind her, ((Galaxy, Lens flare)), short hair, flower field, night sky, cinematic shot. Wallpaper. (Blue color schema), detailed background, a city in the distance"
9
10generator = torch.manual_seed(0)
11image = pipe(prompt, num_inference_steps=25).images[0]
12image.save("./image.png")
