模型:
dreamlike-art/dreamlike-anime-1.0
在您的提示中添加"anime"可使生成的内容更像动漫。在您的提示中添加"photo"可使生成的内容更逼真且具有更好的解剖结构。该模型是基于768x768像素的图片进行训练的,因此请使用768x768像素、704x832像素、832x704像素等。更高的分辨率或非正方形的宽高比可能会产生伪影。
为了获得最佳效果,请在您的提示的开头添加以下内容:
photo anime, masterpiece, high quality, absurdres
使用负面提示可获得最佳效果,例如:
simple background, duplicate, retro style, low quality, lowest quality, 1980s, 1990s, 2000s, 2005 2006 2007 2008 2009 2010 2011 2012 2013, bad anatomy, bad proportions, extra digits, lowres, username, artist name, error, duplicate, watermark, signature, text, extra digit, fewer digits, worst quality, jpeg artifacts, blurry
1girl、girl等会产生略有不同的结果,欢迎尝试并选择您喜欢的效果!
在 dreamlike.art 上免费使用此模型以及 Dreamlike Diffusion 1.0 和 Dreamlike Photoreal 2.0 !
Download dreamlike-anime-1.0.ckpt (2.13GB)
Download dreamlike-anime-1.0.safetensors (2.13GB)
此模型可像其他稳定扩散模型一样使用。有关更多信息,请查看 Stable Diffusion Pipeline 。
from diffusers import StableDiffusionPipeline import torch model_id = "dreamlike-art/dreamlike-anime-1.0" pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16) pipe = pipe.to("cuda") prompt = "anime, masterpiece, high quality, 1girl, solo, long hair, looking at viewer, blush, smile, bangs, blue eyes, skirt, medium breasts, iridescent, gradient, colorful, besides a cottage, in the country" negative_prompt = 'simple background, duplicate, retro style, low quality, lowest quality, 1980s, 1990s, 2000s, 2005 2006 2007 2008 2009 2010 2011 2012 2013, bad anatomy, bad proportions, extra digits, lowres, username, artist name, error, duplicate, watermark, signature, text, extra digit, fewer digits, worst quality, jpeg artifacts, blurry' image = pipe(prompt, negative_prompt=negative_prompt).images[0] image.save("./result.jpg")
该模型根据修改后的CreativeML OpenRAIL-M许可证授权。