plat-diffusion / README.md
p1atdev's picture
Update README.md
3ca6c25
|
raw
history blame
No virus
5.5 kB
metadata
license: creativeml-openrail-m
inference: true
thumbnail: https://bit.ly/3Is8sF8
tags:
  - stable-diffusion
  - stable-diffusion-diffusers
  - text-to-image
  - diffusers

Plat Diffusion v1.3.0

Plat Diffusion is a fine-tuned model based on Waifu Diffusion v1.4 Anime Epoch 1 with images generated with niji・journey.

.safetensors file is here.

kl-f8-anime2.ckpt is recommended for VAE.

thumbnail

Recomended Negative Prompt

nsfw, worst quality, low quality, deleted, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, jpeg artifacts, signature, watermark, username, blurry

Samples

  • Autumn 1girl

sample1

masterpiece, best quality, 1girl, brown hair, long hair, green eyes, autumn, cumulonimbus clouds, lighting, blue sky, falling leaves, garden
Negative prompt: nsfw, worst quality, low quality, deleted, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, jpeg artifacts, signature, watermark, username, blurry,
Steps: 20, Sampler: DPM++ 2M Karras, CFG scale: 7
  • Autumn 1boy

sample2

masterpiece, best quality, 1boy, solo, male focus, medium hair, blonde hair, blue eyes, autumn, cumulonimbus clouds, lighting, blue sky, falling leaves, garden
Negative prompt: nsfw, worst quality, low quality, deleted, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, jpeg artifacts, signature, watermark, username, blurry,
Steps: 20, Sampler: DPM++ 2M Karras, CFG scale: 7
  • Abandoned city

sample3

masterpiece, best quality, scenery, abandoned city, buildings, ruins, rust, blue sky, clouds
Negative prompt: nsfw, worst quality, low quality, deleted, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, jpeg artifacts, signature, watermark, username, blurry,
Steps: 20, Sampler: DPM++ 2M Karras, CFG scale: 7
  • Grey hair school uniform 1girl

sample4

masterpiece, best quality, 1girl, upper body, long hair, grey hair, grey eyes, curly hair, twintails, school uniform, black ribbon, light smile
Negative prompt: nsfw, worst quality, low quality, deleted, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, jpeg artifacts, signature, watermark, username, blurry,
Steps: 20, Sampler: DPM++ 2M Karras, CFG scale: 7
  • 1girl with feathers

sample5

masterpiece, best quality, 1girl, angel, silver hair, white wings, hair flower, school uniform, upper body, sunlight, lily, feathers, ultra detailed illustration
Negative prompt: nsfw, worst quality, low quality, deleted, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, jpeg artifacts, signature, watermark, username, blurry,
Steps: 20, Sampler: DPM++ 2M Karras, CFG scale: 7
  • Many girls

sample6

masterpiece, best quality, high quality, 4girls, everyone, beautiful detailed
Negative prompt: nsfw, worst quality, low quality, deleted, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, jpeg artifacts, signature, watermark, username, blurry,
Steps: 20, Sampler: DPM++ 2M Karras, CFG scale: 7

🧨 Diffusers

from diffusers import StableDiffusionPipeline
import torch

model_id = "p1atdev/plat-diffusion"
pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float32) # not working with float16
pipe = pipe.to("cuda")

prompt = "masterpiece, best quality, 1girl, orange feather, blue crystals, blurry foreground"
negative_prompt = "nsfw, worst quality, low quality, deleted, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, jpeg artifacts, signature, watermark, username, blurry"
image = pipe(prompt, negative_prompt=negative_prompt).images[0]  
    
image.save("girl.png")

License

This model is open access and available to all, with a CreativeML OpenRAIL-M license further specifying rights and usage. The CreativeML OpenRAIL License specifies:

  1. You can't use the model to deliberately produce nor share illegal or harmful outputs or content
  2. The authors claims no rights on the outputs you generate, you are free to use them and are accountable for their use which must not go against the provisions set in the license
  3. You may re-distribute the weights and use the model commercially and/or as a service. If you do, please be aware you have to include the same use restrictions as the ones in the license and share a copy of the CreativeML OpenRAIL-M to all your users (please read the license entirely and carefully) Please read the full license here