|
--- |
|
pipeline_tag: any-to-any |
|
library_name: transformers |
|
tags: |
|
- text-to-image |
|
- image-editing |
|
- image-understanding |
|
- vision-language |
|
- multimodal |
|
- autoregressive |
|
- unified-model |
|
license: mit |
|
--- |
|
|
|
## 🌌 UniPic2-Metaquery-GRPO-Flash |
|
|
|
<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-Metaquery-GRPO-Flash** is a quantized variant of UniPic2-MetaQuery-GRPO, offering end-to-end image understanding, text-to-image (T2I) generation, and image editing. Optimized for efficiency, it runs smoothly on NVIDIA RTX 40-series GPUs with under 16 GB VRAM — without any performance degradation. |
|
<div align="center"> <img src="teaser.png" alt="Model Teaser" width="720"> </div> |
|
<div align="center"> <img src="understanding.png" alt="Model Teaser" width="720"> </div> |
|
|
|
|
|
## 📊 Benchmarks |
|
|
|
<div align="center"> <img src="eval.png" alt="Model Eval" width="720"> </div> |
|
|
|
|
|
## 🧠 Usage |
|
|
|
### 1. Clone the Repository |
|
|
|
```bash |
|
git clone https://github.com/SkyworkAI/UniPic |
|
cd UniPic-2 |
|
``` |
|
|
|
### 2. Set Up the Environment |
|
```bash |
|
conda create -n unipic python=3.10 |
|
conda activate unipic |
|
pip install -r requirements.txt |
|
``` |
|
|
|
|
|
### 3.Text-to-Image Generation |
|
```bash |
|
import torch |
|
from PIL import Image |
|
from unipicv2.pipeline_stable_diffusion_3_kontext import StableDiffusion3KontextPipeline |
|
from unipicv2.transformer_sd3_kontext import SD3Transformer2DKontextModel |
|
from unipicv2.stable_diffusion_3_conditioner import StableDiffusion3Conditioner |
|
from transformers import Qwen2_5_VLForConditionalGeneration, Qwen2_5_VLProcessor |
|
from diffusers import FlowMatchEulerDiscreteScheduler, AutoencoderKL,BitsAndBytesConfig |
|
|
|
# Load model components |
|
pretrained_model_name_or_path = "/path/to/UniPic2-Metaquery-Flash/UniPic2-Metaquery" |
|
vlm_path = "/path/to/UniPic2-Metaquery-Flash/Qwen2.5-VL-7B-Instruct-AWQ" |
|
|
|
quant = "int4" # {"int4", "fp16"} |
|
|
|
bnb4 = BitsAndBytesConfig( |
|
load_in_4bit=True, |
|
bnb_4bit_use_double_quant=True, |
|
bnb_4bit_quant_type="nf4", |
|
bnb_4bit_compute_dtype=torch.float16, # 与 LMM/Cond 对齐 |
|
) |
|
|
|
if quant == "int4": |
|
transformer = SD3Transformer2DKontextModel.from_pretrained( |
|
pretrained_model_name_or_path, subfolder="transformer", |
|
quantization_config=bnb4, device_map="auto", low_cpu_mem_usage=True |
|
) |
|
elif quant == "fp16": |
|
transformer = SD3Transformer2DKontextModel.from_pretrained( |
|
pretrained_model_name_or_path, subfolder="transformer", |
|
torch_dtype=torch.float16, device_map="auto", low_cpu_mem_usage=True |
|
) |
|
else: |
|
raise ValueError(f"Unsupported quant: {quant}") |
|
|
|
|
|
vae = AutoencoderKL.from_pretrained( |
|
pretrained_model_name_or_path, subfolder="vae", |
|
torch_dtype=torch.float16, device_map="auto", low_cpu_mem_usage=True).cuda() |
|
|
|
# Load Qwen2.5-VL model |
|
lmm = Qwen2_5_VLForConditionalGeneration.from_pretrained( |
|
vlm_path, |
|
torch_dtype=torch.bfloat16,device_map="auto", |
|
attn_implementation="flash_attention_2") |
|
|
|
processor = Qwen2_5_VLProcessor.from_pretrained(vlm_path) |
|
processor.chat_template = processor.chat_template.replace( |
|
"{% if loop.first and message['role'] != 'system' %}<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n{% endif %}", |
|
"") |
|
|
|
# 加上cuda |
|
conditioner = StableDiffusion3Conditioner.from_pretrained( |
|
pretrained_model_name_or_path, subfolder="conditioner", torch_dtype=torch.float16).cuda() |
|
|
|
scheduler = FlowMatchEulerDiscreteScheduler.from_pretrained(pretrained_model_name_or_path, subfolder="scheduler") |
|
|
|
# Create pipeline (note: text encoders set to None) |
|
pipeline = StableDiffusion3KontextPipeline( |
|
transformer=transformer, vae=vae, |
|
text_encoder=None, tokenizer=None, |
|
text_encoder_2=None, tokenizer_2=None, |
|
text_encoder_3=None, tokenizer_3=None, |
|
scheduler=scheduler) |
|
|
|
# Prepare prompts |
|
prompt = 'a pig with wings and a top hat flying over a happy futuristic scifi city' |
|
negative_prompt = '' |
|
|
|
messages = [[{"role": "user", "content": [{"type": "text", "text": f'Generate an image: {txt}'}]}] |
|
for txt in [prompt, negative_prompt]] |
|
|
|
texts = [processor.apply_chat_template(msg, tokenize=False, add_generation_prompt=True) for msg in messages] |
|
inputs = processor(text=texts, images=None, videos=None, padding=True, return_tensors="pt").to("cuda") |
|
|
|
# Process with Qwen2.5-VL |
|
input_ids, attention_mask = inputs.input_ids, inputs.attention_mask |
|
input_ids = torch.cat([input_ids, input_ids.new_zeros(2, conditioner.config.num_queries)], dim=1) |
|
attention_mask = torch.cat([attention_mask, attention_mask.new_ones(2, conditioner.config.num_queries)], dim=1) |
|
inputs_embeds = lmm.get_input_embeddings()(input_ids) |
|
inputs_embeds[:, -conditioner.config.num_queries:] = conditioner.meta_queries[None].expand(2, -1, -1) |
|
|
|
outputs = lmm.model(inputs_embeds=inputs_embeds, attention_mask=attention_mask, use_cache=False) |
|
hidden_states = outputs.last_hidden_state[:, -conditioner.config.num_queries:] |
|
prompt_embeds, pooled_prompt_embeds = conditioner(hidden_states) |
|
|
|
# Generate image |
|
image = pipeline( |
|
prompt_embeds=prompt_embeds[:1], |
|
pooled_prompt_embeds=pooled_prompt_embeds[:1], |
|
negative_prompt_embeds=prompt_embeds[1:], |
|
negative_pooled_prompt_embeds=pooled_prompt_embeds[1:], |
|
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") |
|
print(f"Image saved to text2image.png (quant={quant})") |
|
``` |
|
|
|
|
|
### 4. Image Editing |
|
```bash |
|
# Load image for editing |
|
image = Image.open("text2image.png") |
|
image = fix_longer_edge(image, image_size=512) |
|
|
|
prompt = "remove the pig's hat" |
|
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." |
|
|
|
# Prepare messages with image input |
|
messages = [[{"role": "user", "content": [{"type": "image", "image": image}, {"type": "text", "text": txt}]}] |
|
for txt in [prompt, negative_prompt]] |
|
|
|
texts = [processor.apply_chat_template(msg, tokenize=False, add_generation_prompt=True) for msg in messages] |
|
|
|
min_pixels = max_pixels = int(image.height * 28 / 32 * image.width * 28 / 32) |
|
inputs = processor( |
|
text=texts, images=[image]*2, |
|
min_pixels=min_pixels, max_pixels=max_pixels, |
|
videos=None, padding=True, return_tensors="pt").cuda() |
|
|
|
# Process with vision understanding |
|
input_ids, attention_mask, pixel_values, image_grid_thw = \ |
|
inputs.input_ids, inputs.attention_mask, inputs.pixel_values, inputs.image_grid_thw |
|
|
|
input_ids = torch.cat([input_ids, input_ids.new_zeros(2, conditioner.config.num_queries)], dim=1) |
|
attention_mask = torch.cat([attention_mask, attention_mask.new_ones(2, conditioner.config.num_queries)], dim=1) |
|
inputs_embeds = lmm.get_input_embeddings()(input_ids) |
|
inputs_embeds[:, -conditioner.config.num_queries:] = conditioner.meta_queries[None].expand(2, -1, -1) |
|
|
|
image_embeds = lmm.visual(pixel_values, grid_thw=image_grid_thw) |
|
image_token_id = processor.tokenizer.convert_tokens_to_ids('<|image_pad|>') |
|
inputs_embeds[input_ids == image_token_id] = image_embeds |
|
|
|
lmm.model.rope_deltas = None |
|
outputs = lmm.model(inputs_embeds=inputs_embeds, attention_mask=attention_mask, |
|
image_grid_thw=image_grid_thw, use_cache=False) |
|
|
|
hidden_states = outputs.last_hidden_state[:, -conditioner.config.num_queries:] |
|
prompt_embeds, pooled_prompt_embeds = conditioner(hidden_states) |
|
|
|
# Generate edited image |
|
edited_image = pipeline( |
|
image=image, |
|
prompt_embeds=prompt_embeds[:1], |
|
pooled_prompt_embeds=pooled_prompt_embeds[:1], |
|
negative_prompt_embeds=prompt_embeds[1:], |
|
negative_pooled_prompt_embeds=pooled_prompt_embeds[1:], |
|
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_image.png") |
|
print(f"Image saved to edited_image.png (quant={quant})") |
|
|
|
``` |
|
|
|
|
|
## 📄 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}, |
|
} |
|
``` |