Instructions to use hongxingli/P2R-2B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use hongxingli/P2R-2B with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="hongxingli/P2R-2B")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("hongxingli/P2R-2B")
model = AutoModelForMultimodalLM.from_pretrained("hongxingli/P2R-2B")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use hongxingli/P2R-2B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "hongxingli/P2R-2B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "hongxingli/P2R-2B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/hongxingli/P2R-2B

SGLang

How to use hongxingli/P2R-2B with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "hongxingli/P2R-2B" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "hongxingli/P2R-2B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "hongxingli/P2R-2B" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "hongxingli/P2R-2B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use hongxingli/P2R-2B with Docker Model Runner:
```
docker model run hf.co/hongxingli/P2R-2B
```

P2R-2B

This repository contains the P2R-2B, introduced in Perceive-to-Reason: Decoupling Perception and Reasoning for Fine-Grained Visual Reasoning.

Model Description

P2R-2B is a fine-grained visual reasoning model built upon Qwen3-VL-2B-Instruct. It performs inference under the P2R framework, a two-stage visual reasoning framework that decouples perception from reasoning. Training is powered by PRA-GRPO, a role-aware alternating RL strategy.

Model Performance

Model	V-Star	HR-Bench-4K	HR-Bench-8K	MME-RealWorld-Lite
Qwen3-VL-Instruct-2B	74.9	70.4	64.6	47.3
P2R-2B	84.3	75.1	74.8	51.3
Δ	+9.4	+4.7	+10.2	+4.0

Usage

from transformers import AutoProcessor, Qwen3VLForConditionalGeneration

model = Qwen3VLForConditionalGeneration.from_pretrained("hongxingli/P2R-2B")
processor = AutoProcessor.from_pretrained("hongxingli/P2R-2B")

For the full two-stage P2R inference pipeline, please refer to our code repository.

Citation

@misc{li2026perceivetoreasondecouplingperceptionreasoning,
      title={Perceive-to-Reason: Decoupling Perception and Reasoning for Fine-Grained Visual Reasoning}, 
      author={Hongxing Li and Xiufeng Huang and Dingming Li and Wenjing Jiang and Zixuan Wang and Haolei Xu and Hanrong Zhang and Haiwen Hong and Longtao Huang and Hui Xue and Weiming Lu and Jun Xiao and Yueting Zhuang and Yongliang Shen},
      year={2026},
      eprint={2607.01191},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2607.01191}, 
}