---
license: apache-2.0
datasets:
- FreedomIntelligence/PubMedVision
language:
- en
- zh
pipeline_tag: text-generation
tags:
  - vision
  - image-text-to-text
---
<div align="center">
<h1>
  HuatuoGPT-Vision-7B
</h1>
</div>

<div align="center">
<a href="https://github.com/FreedomIntelligence/HuatuoGPT-Vision" target="_blank">GitHub</a> | <a href="https://arxiv.org/abs/2406.19280" target="_blank">Paper</a>
</div>

## Introduction
We convert HuatuoGPT-Vision into Huggingface LLaVA format, so you can run the model using VLLM or other frameworks. The original model can be found here: [HuatuoGPT-Vision-7B](https://huggingface.co/FreedomIntelligence/HuatuoGPT-Vision-7B).

# Quick Start

### 1. Deploy the model using [VLLM](https://github.com/vllm-project/vllm/tree/main)
```bash
python -m vllm.entrypoints.openai.api_server \
--model huatuogpt_vision_model_path  \
--tensor_parallel_size 1 \
--gpu_memory_utilization 0.8 \
--served-model-name huatuogpt_vision_7b \
--chat-template "{%- if messages[0]['role'] == 'system' -%}\n    {%- set system_message = messages[0]['content'] -%}\n    {%- set messages = messages[1:] -%}\n{%- else -%}\n    {% set system_message = '' -%}\n{%- endif -%}\n\n{%- for message in messages -%}\n    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}\n        {{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}\n    {%- endif -%}\n\n    {%- if message['role'] == 'user' -%}\n        {{ '<|user|>\n' + message['content'] + '\n' }}\n    {%- elif message['role'] == 'assistant' -%}\n        {{ '<|assistant|>\n' + message['content'] + '\n' }}\n    {%- endif -%}\n{%- endfor -%}\n\n{%- if add_generation_prompt -%}\n    {{ '<|assistant|>' }}\n{% endif %}" \
--port 9559 --max-model-len 2048 > vllm_openai_server.log 2>&1 &
```

### 2. Model inference
```python
from openai import OpenAI
from PIL import Image
import base64
import io

def get_image(image_path):
    image = Image.open(image_path).convert('RGB')
    img_type = image.format
    if not img_type:
        img_type = image_path.split('.')[-1]
    byte_arr = io.BytesIO()
    image.save(byte_arr, format=img_type)
    byte_arr.seek(0)
    image = base64.b64encode(byte_arr.getvalue()).decode()
    return image, img_type


client = OpenAI(
    base_url="http://localhost:9559/v1",
    api_key="token-abc123"
)
image_path = 'your_image_path'
image, img_type = get_image(image_path)


inputcontent = [{
    "type": "text",
    "text": '<image>\nWhat does the picture show?'
}]

inputcontent.append({
    "type": "image_url",
    "image_url": {
        "url": f"data:image/{img_type};base64,{image}"
    }
})

response = client.chat.completions.create(
    model="huatuogpt_vision_7b",
    messages=[
        {"role": "user", "content": inputcontent}
    ],
    temperature=0.2
)
print(response.choices[0].message.content)
```

# <span id="Start">Citation</span>

```
@misc{chen2024huatuogptvisioninjectingmedicalvisual,
      title={HuatuoGPT-Vision, Towards Injecting Medical Visual Knowledge into Multimodal LLMs at Scale}, 
      author={Junying Chen and Ruyi Ouyang and Anningzhe Gao and Shunian Chen and Guiming Hardy Chen and Xidong Wang and Ruifei Zhang and Zhenyang Cai and Ke Ji and Guangjun Yu and Xiang Wan and Benyou Wang},
      year={2024},
      eprint={2406.19280},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2406.19280}, 
}
```