---
title: qwen2-0.5B-functioncall
canonical_url: "https://www.modelscope.cn/models/longdream/qwen2-0.5B-functioncall"
md_url: "https://www.modelscope.cn/models/longdream/qwen2-0.5B-functioncall.md"
repository: longdream/qwen2-0.5B-functioncall
chinese_name: "通义千问2-0.5BF-functioncall微调"
last_updated: 2024-07-22
license: "Apache License 2.0"
pipeline_tag: text-generation
tasks:
  - text-generation
model_type:
  - qwen2
architectures:
  - Qwen2ForCausalLM
parameters: 494.0M
tensor_type:
  - BF16
library_name:
  - safetensors
  - pytorch
frameworks:
  - Pytorch
inference_backends:
  - "deploy_task text/emb"
  - "lmdeploy 0.9.1"
  - "lmdeploy_turbomind 0.9.1"
  - "sglang 0.5.2"
  - "vllm 0.9.2"
downloads: 145
stars: 3
---

# qwen2-0.5B-functioncall

> qwen2-0.5B-functioncall - longdream 在 ModelScope 开源的模型。qwen2-0.5B fucntion call 简单微调版本

longdream/qwen2-0.5B-functioncall 是 ModelScope 魔搭社区上的 494.0M 参数text-generation模型，采用 Apache License 2.0 许可，可用 deploy_task text/emb、lmdeploy 0.9.1、lmdeploy_turbomind 0.9.1 部署。

- **Repository**: longdream/qwen2-0.5B-functioncall
- **License**: Apache License 2.0
- **Tasks**: text-generation
- **Parameters**: 494.0M
- **Inference backends**: deploy_task text/emb, lmdeploy 0.9.1, lmdeploy_turbomind 0.9.1, sglang 0.5.2, vllm 0.9.2
- **Downloads**: 145
- **Stars**: 3
- **Last updated**: 2024-07-22

Source: https://www.modelscope.cn/models/longdream/qwen2-0.5B-functioncall

---

# 模型介绍  
## 模型概述 我们基于 [qwen2 0.5B](https://www.modelscope.com/models/qwen2-0.5B) 模型进行了微调，使其具备了 function call 功能。下面示例使用了llamafactory 运行模型来提供api进行测试


### 您可以通过如下git clone命令，或者ModelScope SDK来下载模型
#### 您可以 
SDK下载
```bash
#安装ModelScope
pip  install  modelscope
```
```python
#SDK模型下载
from modelscope import snapshot_download
model_dir = snapshot_download('longdream/qwen2-0.5B-functioncall')
```
Git下载
```
#Git模型下载
git clone https://www.modelscope.cn/longdream/qwen2-0.5B-functioncall.git
```
```python
#fucntion call 进行测试
import os
import json
from openai import OpenAI
from typing import Sequence
import requests

os.environ["OPENAI_BASE_URL"] = "http://192.168.1.103:38000/v1"
os.environ["OPENAI_API_KEY"] = "0"


def get_rubbish_category(keyword):
    url = f"https://api.timelessq.com/garbage?keyword={keyword}"
    response = requests.request("GET", url)
    output_str_list = []
    for item in response.json()['data']:
        output_str_list.append(f"{item['name']}: {item['categroy']}")
    return '\n'.join(output_str_list)


def get_song_information(keyword):
    url = f"https://api.timelessq.com/music/tencent/search?keyword={keyword}"
    response = requests.request("GET", url)
    song_infor = response.json()['data']['list'][0]
    singer = '' if not song_infor['singer'] else song_infor['singer'][0]['name']
    return f"歌曲: {keyword}\n歌手: {singer}\n时长: {song_infor['interval']}秒\n专辑名称: {song_infor['albumname']}"


def get_cartoon_information(title):
    url = f"https://api.timelessq.com/bangumi?title={title}"
    response = requests.request("GET", url)
    data = response.json()['data'][0]
    return f"标题: {data['title']}\n类型：{data['type']}\n语言：{data['lang']}\n出品方：{data['officialSite']}\n上映时间：{data['begin']}\n完结事件：{data['end']}"


tool_map = {"get_rubbish_category": get_rubbish_category,
            "get_song_information": get_song_information,
            "get_cartoon_information": get_cartoon_information}


if __name__ == "__main__":
    client = OpenAI()
    tools = [
        {
            "type": "function",
            "function": {
                            "name": "get_cartoon_information",
                            "description": "根据用户提供的动漫标题，查询该动漫的相关信息。",
                            "parameters": {
                                "type": "object",
                                "properties": {
                                    "title": {
                                        "type": "string",
                                        "description": "动漫",
                                    },
                                },
                                "required": ["title"],
                            }
                        }
        },
        {
            "type": "function",
            "function": {
                            "name": "get_song_information",
                            "description": "根据用户提供的歌曲名称，查询歌曲相关信息，包括歌手、时长、专辑名称等。",
                            "parameters": {
                                "type": "object",
                                "properties": {
                                    "keyword": {
                                        "type": "string",
                                        "description": "歌曲名称",
                                    },
                                },
                                "required": ["keyword"],
                            }
                        }
        }
    ]

    messages = []
    messages.append({"role": "system", "content": "你是一个有用的小助手，请调用下面的工具来回答用户的问题，参考工具输出进行回答。"})
    # messages.append({"role": "user", "content": "鸡蛋壳属于哪种类型的垃圾？"})
    # messages.append({"role": "user", "content": "爱在西元前是谁唱的，来自哪张专辑？"})
    messages.append({"role": "user", "content": "歌曲《七里香》是哪个国家的，歌手是谁？"})
    result = client.chat.completions.create(messages=messages, model="Qwen2-0.5B-agent", tools=tools)
    print(result)
    tool_call = result.choices[0].message.tool_calls[0].function
    print(tool_call)
    name, arguments = tool_call.name, json.loads(tool_call.arguments)
    messages.append({"role": "function", "content": json.dumps({"name": name, "argument": arguments}, ensure_ascii=False)})
    tool_result = tool_map[name](**arguments)
    messages.append({"role": "tool", "content": "工具输出结果为: " + tool_result})
    for msg in messages:
        print('--->', msg)
    result = client.chat.completions.create(messages=messages, model="Qwen2-0.5B-agent")
    print("Answer: ", result.choices[0].message.content)
```
