UniPic2-SD3.5M-Kontext-2B

🌌 UniPic2-SD3.5M-Kontext-2B

<div align="center"> <img src="skywork-logo.png" alt="Skywork Logo" width="500"> </div> <p align="center"> <a href="https://github.com/SkyworkAI/UniPic"> <img src="https://img.shields.io/badge/GitHub-UniPic-blue?logo=github" alt="GitHub Repo"> </a> <a href="https://github.com/SkyworkAI/UniPic/stargazers"> <img src="https://img.shields.io/github/stars/SkyworkAI/UniPic?style=social" alt="GitHub Stars"> </a> <a href="https://github.com/SkyworkAI/UniPic/network/members"> <img src="https://img.shields.io/github/forks/SkyworkAI/UniPic?style=social" alt="GitHub Forks"> </a> </p>

📖 Introduction

UniPic2-SD3.5M-Kontext-2B is a post-trained **T2I ** model built on the SD3.5-Medium. It focuses on text-to-image generation and image editing, delivering strong quality with a fast generation speed. It runs smoothly on a single 16 GB consumer GPU.

<div align="center"> <img src="teaser.png" alt="Model Teaser" width="720"> </div>

📊 Benchmarks

UniPic2-SD3.5M-Kontext-2B w/o GRPO achieves competitive results across a variety of vision-language tasks:

🧠 Usage

1. Clone the Repository

git clone https://github.com/SkyworkAI/UniPic
cd UniPic-2

2. Set Up the Environment

conda create -n unipic python=3.10
conda activate unipic
pip install -r requirements.txt

3.Text-to-Image Generation

import torch
from PIL import Image
from unipicv2.pipeline_stable_diffusion_3_kontext import StableDiffusion3KontextPipeline
from unipicv2.transformer_sd3_kontext import SD3Transformer2DKontextModel
from diffusers import FlowMatchEulerDiscreteScheduler, AutoencoderKL
from transformers import CLIPTextModelWithProjection, CLIPTokenizer, T5EncoderModel, T5TokenizerFast

# Load model components
pretrained_model_name_or_path = "Skywork/UniPic2-SD3.5M-Kontext-2B"

transformer = SD3Transformer2DKontextModel.from_pretrained(
        pretrained_model_name_or_path, subfolder="transformer", torch_dtype=torch.bfloat16).cuda()

vae = AutoencoderKL.from_pretrained(
    pretrained_model_name_or_path, subfolder="vae",
    torch_dtype=torch.bfloat16, device_map="auto", low_cpu_mem_usage=True
).cuda()

# Load text encoders
text_encoder = CLIPTextModelWithProjection.from_pretrained(
    pretrained_model_name_or_path, subfolder="text_encoder", torch_dtype=torch.bfloat16, device_map="auto", low_cpu_mem_usage=True
).cuda()
tokenizer = CLIPTokenizer.from_pretrained(pretrained_model_name_or_path, subfolder="tokenizer")

text_encoder_2 = CLIPTextModelWithProjection.from_pretrained(
    pretrained_model_name_or_path, subfolder="text_encoder_2", torch_dtype=torch.bfloat16, device_map="auto", low_cpu_mem_usage=True
).cuda()
tokenizer_2 = CLIPTokenizer.from_pretrained(pretrained_model_name_or_path, subfolder="tokenizer_2")

text_encoder_3 = T5EncoderModel.from_pretrained(
    pretrained_model_name_or_path, subfolder="text_encoder_3", torch_dtype=torch.bfloat16, device_map="auto", low_cpu_mem_usage=True
).cuda()
tokenizer_3 = T5TokenizerFast.from_pretrained(pretrained_model_name_or_path, subfolder="tokenizer_3")

scheduler = FlowMatchEulerDiscreteScheduler.from_pretrained(
    pretrained_model_name_or_path, subfolder="scheduler"
)

# Create pipeline
pipeline = StableDiffusion3KontextPipeline(
    transformer=transformer, vae=vae,
    text_encoder=text_encoder, tokenizer=tokenizer,
    text_encoder_2=text_encoder_2, tokenizer_2=tokenizer_2,
    text_encoder_3=text_encoder_3, tokenizer_3=tokenizer_3,
    scheduler=scheduler)

# Generate image
image = pipeline(
    prompt='a pig with wings and a top hat flying over a happy futuristic scifi city',
    negative_prompt='blurry, low quality, low resolution, distorted, deformed, broken content, missing parts, damaged details, artifacts, glitch, noise, pixelated, grainy, compression artifacts, bad composition, wrong proportion, incomplete editing, unfinished, unedited areas.',
    height=512, width=384,
    num_inference_steps=50,
    guidance_scale=3.5,
    generator=torch.Generator(device=transformer.device).manual_seed(42)
).images[0]

image.save("text2image.png")

4. Image Editing

# Load and preprocess image
def fix_longer_edge(x, image_size, factor=32):
    w, h = x.size
    if w >= h:
        target_w = image_size
        target_h = h * (target_w / w)
        target_h = round(target_h / factor) * factor
    else:
        target_h = image_size
        target_w = w * (target_h / h)
        target_w = round(target_w / factor) * factor
    x = x.resize(size=(target_w, target_h))
    return x

image = Image.open("text2image.png")
image = fix_longer_edge(image, image_size=512)

negative_prompt = "blurry, low quality, low resolution, distorted, deformed, broken content, missing parts, damaged details, artifacts, glitch, noise, pixelated, grainy, compression artifacts, bad composition, wrong proportion, incomplete editing, unfinished, unedited areas."

# Edit image
edited_image = pipeline(
    image=image,
    prompt="remove the pig's hat",
    negative_prompt=negative_prompt,
    height=image.height, width=image.width,
    num_inference_steps=50,
    guidance_scale=3.5,
    generator=torch.Generator(device=transformer.device).manual_seed(42)
).images[0]

edited_image.save("edited_img.png")

📄 License

This model is released under the MIT License.

Citation

If you use Skywork-UniPic in your research, please cite:

@misc{wang2025skyworkunipicunifiedautoregressive,
      title={Skywork UniPic: Unified Autoregressive Modeling for Visual Understanding and Generation}, 
      author={Peiyu Wang and Yi Peng and Yimeng Gan and Liang Hu and Tianyidan Xie and Xiaokun Wang and Yichen Wei and Chuanxin Tang and Bo Zhu and Changshi Li and Hongyang Wei and Eric Li and Xuchen Song and Yang Liu and Yahui Zhou},
      year={2025},
      eprint={2508.03320},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2508.03320}, 
}

@misc{wei2025skyworkunipic20building,
      title={Skywork UniPic 2.0: Building Kontext Model with Online RL for Unified Multimodal Model}, 
      author={Hongyang Wei and Baixin Xu and Hongbo Liu and Cyrus Wu and Jie Liu and Yi Peng and Peiyu Wang and Zexiang Liu and Jingwen He and Yidan Xietian and Chuanxin Tang and Zidong Wang and Yichen Wei and Liang Hu and Boyi Jiang and William Li and Ying He and Yang Liu and Xuchen Song and Eric Li and Yahui Zhou},
      year={2025},
      eprint={2509.04548},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2509.04548}, 
}