Diffusers
from diffusers import StableDiffusionPipeline
import torch
model_id = "acondess/lineart"
pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
pipe = pipe.to("cuda")
prompt = "a photo of an astronaut riding a horse on mars"
image = pipe(prompt).images[0]
image.save("astronaut_rides_horse.png")
F 复现文档:
https://cloud.alipay.com/developer/0945907cc6974a5bb6efc9660672f5fb/post/212fb91ee59443aa81257d8e70cfe22f
https://www.yuque.com/dengchao-enlft/nngkbu/qlknlkc2pmkceyi3?singleDoc# 《Stable Diffusion基于Dreambooth训练自定义模型》
底模为revAnimated_v122.safetensors
使用Dreambooth训练
数据集为Coloring Book Objects Dataset
提示词:
lineart of cat,coloring book,high quality,
参数:
Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 629188890, Size: 512x512, Model hash: 77e875916e, Model: lineart_2000_lora, Clip skip: 2, ENSD: 31337, Version: v1.5.1