Andyrasika/avatar_diffusion

Dreambooth style: Avatar Dreambooth finetuning of Stable Diffusion (v1.5.1) on Avatar art style.

About This text-to-image stable diffusion model was trained with dreambooth. Put in a text prompt and generate your own Avatar style image!

(Image taken from Lambdalabs repo)

from diffusers import DiffusionPipeline, UniPCMultistepScheduler
import torch
from torch import autocast
pipeline = DiffusionPipeline.from_pretrained(
    "Andyrasika/avatar_diffusion",
    custom_pipeline="lpw_stable_diffusion",

    torch_dtype=torch.float16
)
pipeline.scheduler = UniPCMultistepScheduler.from_config(pipeline.scheduler.config)
pipeline.to("cuda")

pipeline.enable_vae_tiling()
pipeline.enable_xformers_memory_efficient_attention()

prompt = "Yoda, avatarart style"
scale = 7.5
n_samples = 4

with autocast("cuda"):
  images = pipeline(n_samples*[prompt], guidance_scale=scale).images

for idx, im in enumerate(images):
  im.save(f"{idx:06}.png")