svjack/Stable-Diffusion-Pokemon-ja

Japanese Stable Diffusion Pokemon Model Card

Stable-Diffusion-Pokemon-ja is a Japanese-specific latent text-to-image diffusion model capable of generating Pokemon images given any text input.

This model was trained by using a powerful text-to-image model, diffusers For more information about our training method, see train_ja_model.py.

Model Details

• Developed by: Zhipeng Yang
• Model type: Diffusion-based text-to-image generation model
• Language(s): Japanese
• License: The CreativeML OpenRAIL M license is an Open RAIL M license, adapted from the work that BigScience and the RAIL Initiative are jointly carrying in the area of responsible AI licensing. See also the article about the BLOOM Open RAIL license on which our license is based.
• Model Description: This is a model that can be used to generate and modify images based on text prompts. It is a Latent Diffusion Model (LDM) that used Stable Diffusion as a pre-trained model.
• Resources for more information: https://github.com/svjack/Stable-Diffusion-Pokemon

Examples

Firstly, install our package as follows. This package is modified 🤗's Diffusers library to run Japanese Stable Diffusion.

pip install git+https://github.com/rinnakk/japanese-stable-diffusion
sudo apt-get install git-lfs
git clone https://huggingface.co/svjack/Stable-Diffusion-Pokemon-ja

Run this command to log in with your HF Hub token if you haven't before:

huggingface-cli login

Running the pipeline with the LMSDiscreteScheduler scheduler:

from japanese_stable_diffusion import JapaneseStableDiffusionPipeline
import torch

from torch import autocast
from diffusers import LMSDiscreteScheduler

scheduler = LMSDiscreteScheduler(beta_start=0.00085, beta_end=0.012,
     beta_schedule="scaled_linear", num_train_timesteps=1000)

#pretrained_model_name_or_path = "jap_model_26000"

#### sudo apt-get install git-lfs
#### git clone https://huggingface.co/svjack/Stable-Diffusion-Pokemon-ja
pretrained_model_name_or_path = "Stable-Diffusion-Pokemon-ja"

pipe = JapaneseStableDiffusionPipeline.from_pretrained(pretrained_model_name_or_path,
                                                           scheduler=scheduler, use_auth_token=True)

pipe = pipe.to("cuda")

#### disable safety_checker
pipe.safety_checker = lambda images, clip_input: (images, False)

imgs = pipe("鉢植えの植物を頭に載せた漫画のキャラクター",
                    num_inference_steps = 100
)
image = imgs.images[0]
    
image.save("output.png")

Generator Results comparison

https://github.com/svjack/Stable-Diffusion-Pokemon