About ai-forever/ru-dalle
ai-forever/ru-dalle is an open-source project on GitHub, mainly written in Jupyter Notebook. Generate images from texts. In Russian It currently holds 1,642 stars and 242 forks with 0 open issues, and was last pushed on an unknown date (repository created unknown).
Project Overview
AI Homed tracks it on the AI Image Projects board and on the AI AI Image Projects list.
GitHub Repository Details
README
ruDALL-E
Generate images from texts
pip install rudalle==1.1.3
🤗 HF Models:
ruDALL-E Malevich (XL) \ ruDALL-E Emojich (XL) (readme here) \ ruDALL-E Surrealist (XL) \ ruDALL-E Kandinsky (XXL) (soon)Minimal Example:
Example usage ruDALL-E Malevich (XL) with 3.5GB vRAM!
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
seejupyters/ruDALLE-image-prompts-A100.ipynb
text, seed = 'Храм Василия Блаженного', 42
skyes = [red_sky, sunny_sky, cloudy_sky, night_sky]
VideoDALL-E | ruCogVideo by @cene555
Video generation exampleAspect ratio images -->NEW<--
Kandinsky 12B
Request access: Here
роботы акварелью в стиле ван гога

FID = 15.4 (COCO Valid)
🚀 Contributors 🚀
- @bes shared great idea and realization with IDWT for decoding images with higher quality 512x512! 😈💪 thanks a lot for your constructive advices, appreciate it
- @neverix thanks a lot for contributing for speed up of inference
- @Igor Pavlov trained model and prepared code with super-resolution
- @oriBetelgeuse thanks a lot for easy API of generation using image prompt
- @Alex Wortega created first FREE version colab notebook with fine-tuning ruDALL-E Malevich (XL) on sneakers domain 💪
- @Anton Lozhkov Integrated to Huggingface Spaces with Gradio, see here
