Skip to content

zeroshot-ai/ru-dalle

 
 

Repository files navigation

ruDALL-E

Generate images from texts

Apache license Downloads Coverage Status pipeline pre-commit.ci status

pip install rudalle==1.0.0

🤗 HF Models:

ruDALL-E Malevich (XL)
ruDALL-E Emojich (XL) (readme here)

Minimal Example:

Open In Colab Kaggle Hugging Face Spaces

Example usage ruDALL-E Malevich (XL) with 3.5GB vRAM! Open In Colab

Finetuning example Open In Colab

generation by ruDALLE:

import ruclip
from rudalle.pipelines import generate_images, show, super_resolution, cherry_pick_by_ruclip
from rudalle import get_rudalle_model, get_tokenizer, get_vae, get_realesrgan
from rudalle.utils import seed_everything

# prepare models:
device = 'cuda'
dalle = get_rudalle_model('Malevich', pretrained=True, fp16=True, device=device)
tokenizer = get_tokenizer()
vae = get_vae(dwt=True).to(device)

# pipeline utils:
realesrgan = get_realesrgan('x2', device=device)
clip, processor = ruclip.load('ruclip-vit-base-patch32-384', device=device)
clip_predictor = ruclip.Predictor(clip, processor, device, bs=8)
text = 'радуга на фоне ночного города'

seed_everything(42)
pil_images = []
scores = []
for top_k, top_p, images_num in [
    (2048, 0.995, 24),
]:
    _pil_images, _scores = generate_images(text, tokenizer, dalle, vae, top_k=top_k, images_num=images_num, bs=8, top_p=top_p)
    pil_images += _pil_images
    scores += _scores

show(pil_images, 6)

auto cherry-pick by ruCLIP:

top_images, clip_scores = cherry_pick_by_ruclip(pil_images, text, clip_predictor, count=6)
show(top_images, 3)

super resolution:

sr_images = super_resolution(top_images, realesrgan)
show(sr_images, 3)

text, seed = 'красивая тян из аниме', 6955

Image Prompt

see jupyters/ruDALLE-image-prompts-A100.ipynb

text, seed = 'Храм Василия Блаженного', 42
skyes = [red_sky, sunny_sky, cloudy_sky, night_sky]

🚀 Contributors 🚀

Supported by



Social Media

About

ruDALL-E | Generate images from texts

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Jupyter Notebook 98.1%
  • Python 1.9%