### model - 类型:`'safe' | 'nai' | 'furry'` - 默认值:`'nai'` 默认的生成模型。 ### orient - 类型:`'portrait' | 'square' | 'landscape'` - 默认值:`'portrait'` 默认的图片方向。 ### sampler - 类型:`'k_euler_ancestral' | 'k_euler' | 'k_lms' | 'plms' | 'ddim'` -...
There’s a “furry” option for non-human creatures, and its output style is similar to Pixar animation. What NovelAI sacrifices in terms of control, it makes up for in ease of use. When you select the “Generate Images” option, a screen asks you to write out a description of the ...
furry ,pony, pubic hair, mosaic, excrement, faeces, shit, futa, testis" image = pipe(prompt,height=height,width=width,num_inference_steps=steps,guidance_scale=cfg,seed=seed,negative_prompt=negative_prompt).images[0] print('Seed =1993459194', image.argument['seed']) display(image) save_image...
model:Set Model for Generation (safe, nai, furry) orient:Set Image Orientation (portrait, landscape, square) sampler:Set Sampler (k_euler_ancestral, k_euler, k_lms, plms, ddim) anatomy.true:Filter Anatomically Incorrect Images anatomy.false:Allow Anatomically Incorrect Images ...
│ ├── furry_scalie.pt │ ├── furry_transformation.pt │ └── pony.pt ├── LDSR ├── Stable-diffusion │ ├── final-pruned.ckpt -> novelai 的 model.ckpt (pruned) │ ├── final-pruned.vae.pt -> novelai 的 animevae.pt ...
furry_protogen furry_scalie:furry控比较喜欢的 furry_transformation pony:但从这些种子和设置来看,我对“pony”感到惊讶; 附录 另外,根据群友测试和其他地方获取的结果,(),[]是有效的,{}是无效的. ()让关键词权重*1.1 []让关键词权重/1.1 AND关键字可以让两个tag融合,比如white hair and pink hair就会让两...
within Steps: 80, Sampler: DPM2, CFG scale: 7, Seed: 随机, Size: 1280x704(随机), Model hash: 202fcec0, Clip skip: 2, ENSD: 31337 如果你感觉很不错就点个免费的赞吧,让我知道有人在看,之后我也会试着多分享一些内容的。
bad tentacles, poorly drawn tentacles, split tentacles, fused tentacles, missing clit, bad clit, fused clit, colorful clit, black clit, liquid clit, QR code, bar code, censored, safety panties, safety knickers, beard, furry ,pony, pubic hair, mosaic, excrement, faeces, shit, futa, testis ...
Steps: 50, Sampler: Euler a, CFG scale: 12, Seed: 3462281497, Size: 960x640, Model hash: 925997e9 这份TAG生成的图像风格非常一致,大概和hiten_1有关,这也意味着采用的hiten样本中也大多是这些TAG,效果上堪比GAN,而多样性却远胜于GAN,Stable Diffusion真的未来可期。不过这些TAG中不少直接限定了具体的...
import os from IPython.display import clear_output from utils import check_is_model_complete print('正在解压模型') if not check_is_model_complete("./NovelAI_latest_ab21ba3c_paddle"): # 解压 latest 模型 !unzip -o "/home/aistudio/data/data171442/NovelAI_latest_ab21ba3c_paddle.zip" # ...