LoRA text2image fine-tuning - Oedon42/thangka-lora-xl

These are LoRA adaption weights for stabilityai/stable-diffusion-xl-base-1.0. The weights were fine-tuned on the Oedon42/Thang-ka dataset. You can find some example images in the following.

Intended uses & limitations

How to use

from diffusers import DiffusionPipeline

pipe = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0")
pipe.load_lora_weights("Oedon42/thangka-lora-xl")

prompt = "Avalokiteshvara, bodhisattva, Green, Gold, Mountains, Clouds, Lotus throne, Circular halo"
image = pipe(prompt).images[0]

Warning

😰 due to the complexity, the generated thang-ka are in distortion.

We are improving the training, and more detailed Thangka models will be released later.

Notice on the Use of Religious Artworks

1. Prohibition of Commercial Use: These religious artworks are of great religious and cultural significance. Any form of commercial use, including but not limited to using them for advertising, product promotion, or any profit - making activities, is strictly prohibited without prior written permission.

2. Prohibition of Illegal Activities: It is forbidden to modify, distort, or desecrate these artworks in any way that may violate religious beliefs or relevant laws. Activities such as using them for illegal religious propaganda, inciting religious hatred, or any other illegal and criminal acts are strictly prohibited.

3. Legal Consequences: Those who violate the above - mentioned regulations will be held legally accountable. Appropriate legal actions, including claims for compensation for damages, will be taken against violators.

4. Respect for Religious and Cultural Heritage: We encourage everyone to respect the religious and cultural value of these artworks and use them in a way that conforms to laws, regulations, and moral standards.