本文介绍如何通过Ollama结合Qwen2，搭建OpenAI格式的聊天API，并与外部函数结合来拓展模型的更多功能。

tools是OpenAI的Chat Completion API中的一个可选参数，可用于提供函数调用规范（function specifications）。这样做的目的是使模型能够生成符合所提供的规范的函数参数格式。同时，API 实际上不会执行任何函数调用。开发人员需要使用模型输出来执行函数调用。

Ollama支持OpenAI格式API的tool参数，在tool参数中，如果functions提供了参数，Qwen将会决定何时调用什么样的函数，不过Ollama目前还不支持强制使用特定函数的参数tool_choice。

注：本文测试用例参考OpenAI cookbook：https://cookbook.openai.com/examples/how_to_call_functions_with_chat_models

本文主要包含以下三个部分：

• 模型部署：使用Ollama和千问，通过设置template，部署支持Function call的聊天API接口。
• 生成函数参数：指定一组函数并使用 API 生成函数参数。
• 调用具有模型生成的参数的函数：通过实际执行具有模型生成的参数的函数来闭合循环。

01、模型部署

单模型文件下载

使用ModelScope命令行工具下载单个模型，本文使用Qwen2-7B的GGUF格式：

modelscope download --model=qwen/Qwen2-7B-Instruct-GGUF --local_dir . qwen2-7b-instruct-q5_k_m.gguf

Linux环境使用

Liunx用户可使用魔搭镜像环境安装【推荐】

modelscope download --model=modelscope/ollama-linux --local_dir ./ollama-linux
cd ollama-linux
sudo chmod 777 ./ollama-modelscope-install.sh
./ollama-modelscope-install.sh

启动Ollama服务

ollama serve

创建ModelFile

复制模型路径，创建名为“ModelFile”的meta文件，其中设置template，使之支持function call，内容如下：

FROM /mnt/workspace/qwen2-7b-instruct-q5_k_m.gguf


# set the temperature to 0.7 [higher is more creative, lower is more coherent]
PARAMETER temperature 0.7
PARAMETER top_p 0.8
PARAMETER repeat_penalty 1.05
TEMPLATE """{{ if .Messages }}
{{- if or .System .Tools }}<|im_start|>system
{{ .System }}
{{- if .Tools }}


# Tools


You are provided with function signatures within <tools></tools> XML tags. You may call one or more functions to assist with the user query. Don't make assumptions about what values to plug into functions. Here are the available tools:
<tools>{{- range .Tools }}{{ .Function }}{{- end }}</tools>


For each function call, return a JSON object with function name and arguments within <tool_call></tool_call> XML tags as follows:
<tool_call>
{"name": <function-name>, "arguments": <args-json-object>}
</tool_call>{{- end }}<|im_end|>{{- end }}
{{- range .Messages }}
{{- if eq .Role "user" }}
<|im_start|>{{ .Role }}
{{ .Content }}<|im_end|>
{{- else if eq .Role "assistant" }}
<|im_start|>{{ .Role }}
{{- if .Content }}
{{ .Content }}
{{- end }}
{{- if .ToolCalls }}
<tool_call>
{{ range .ToolCalls }}{"name": "{{ .Function.Name }}", "arguments": {{ .Function.Arguments }}}
{{ end }}</tool_call>
{{- end }}<|im_end|>
{{- else if eq .Role "tool" }}
<|im_start|>user
<tool_response>
{{ .Content }}
</tool_response><|im_end|>
{{- end }}
{{- end }}
<|im_start|>assistant
{{ else }}{{ if .System }}<|im_start|>system
{{ .System }}<|im_end|>
{{ end }}{{ if .Prompt }}<|im_start|>user
{{ .Prompt }}<|im_end|>
{{ end }}<|im_start|>assistant
{{ end }}
"""

创建自定义模型

使用ollama create命令创建自定义模型

ollama create myqwen2 --file ./ModelFile

运行模型：

ollama run myqwen2

02、生成函数参数

安装依赖

!pip install scipy --quiet
!pip install tenacity --quiet
!pip install tiktoken --quiet
!pip install termcolor --quiet
!pip install openai --quiet

使用OpenAI的API格式调用本地部署的qwen2模型

import json
import openai
from tenacity import retry, wait_random_exponential, stop_after_attempt
from termcolor import colored  


MODEL = "myqwen2"
client = openai.OpenAI(
    base_url="http://127.0.0.1:11434/v1",
    api_key = "None"
)

实用工具

首先，让我们定义一些实用工具，用于调用聊天完成 API 以及维护和跟踪对话状态。

@retry(wait=wait_random_exponential(multiplier=1, max=40), stop=stop_after_attempt(3))
def chat_completion_request(messages, tools=None, tool_choice=None, model=MODEL):
    try:
        response = client.chat.completions.create(
            model=model,
            messages=messages,
            tools=tools,
            tool_choice=tool_choice,
        )
        return response
    except Exception as e:
        print("Unable to generate ChatCompletion response")
        print(f"Exception: {e}")
        return e

def pretty_print_conversation(messages):
    role_to_color = {
        "system": "red",
        "user": "green",
        "assistant": "blue",
        "function": "magenta",
    }


    for message in messages:
        if message["role"] == "system":
            print(colored(f"system: {message['content']}\n", role_to_color[message["role"]]))
        elif message["role"] == "user":
            print(colored(f"user: {message['content']}\n", role_to_color[message["role"]]))
        elif message["role"] == "assistant" and message.get("function_call"):
            print(colored(f"assistant: {message['function_call']}\n", role_to_color[message["role"]]))
        elif message["role"] == "assistant" and not message.get("function_call"):
            print(colored(f"assistant: {message['content']}\n", role_to_color[message["role"]]))
        elif message["role"] == "function":
            print(colored(f"function ({message['name']}): {message['content']}\n", role_to_color[message["role"]]))

基本概念（https://cookbook.openai.com/examples/how_to_call_functions_with_chat_models#basic-concepts）

这里假设了一个天气 API，并设置了一些函数规范和它进行交互。将这些函数规范传递给 Chat API，以便模型可以生成符合规范的函数参数。

tools = [
    {
        "type": "function",
        "function": {
            "name": "get_current_weather",
            "description": "Get the current weather",
            "parameters": {
                "type": "object",
                "properties": {
                    "location": {
                        "type": "string",
                        "description": "The city and state, e.g. San Francisco, CA",
                    },
                    "format": {
                        "type": "string",
                        "enum": ["celsius", "fahrenheit"],
                        "description": "The temperature unit to use. Infer this from the users location.",
                    },
                },
                "required": ["location", "format"],
            },
        }
    },
    {
        "type": "function",
        "function": {
            "name": "get_n_day_weather_forecast",
            "description": "Get an N-day weather forecast",
            "parameters": {
                "type": "object",
                "properties": {
                    "location": {
                        "type": "string",
                        "description": "The city and state, e.g. San Francisco, CA",
                    },
                    "format": {
                        "type": "string",
                        "enum": ["celsius", "fahrenheit"],
                        "description": "The temperature unit to use. Infer this from the users location.",
                    },
                    "num_days": {
                        "type": "integer",
                        "description": "The number of days to forecast",
                    }
                },
                "required": ["location", "format", "num_days"]
            },
        }
    },
]

如果我们向模型询问当前的天气情况，它将会反问，希望获取到进一步的更多的参数信息。

messages = []
messages.append({"role": "system", "content": "Don't make assumptions about what values to plug into functions. Ask for clarification if a user request is ambiguous."})
messages.append({"role": "user", "content": "hi ，can you tell me what's the weather like today"})
chat_response = chat_completion_request(
    messages, tools=tools
)
assistant_message = chat_response.choices[0].message
messages.append(assistant_message)
assistant_message

ChatCompletionMessage(content='Of course, I can help with that. To provide accurate information, could you please specify the city and state you are interested in?', role='assistant', function_call=None, tool_calls=None)

一旦我们通过对话提供缺失的参数信息，模型就会为我们生成适当的函数参数。

messages.append({"role": "user", "content": "I'm in Glasgow, Scotland."})
chat_response = chat_completion_request(
    messages, tools=tools
)
assistant_message = chat_response.choices[0].message
messages.append(assistant_message)
assistant_message

ChatCompletionMessage(content='', role='assistant', function_call=None, tool_calls=[ChatCompletionMessageToolCall(id='call_qq8e5z9w', function=Function(arguments='{"location":"Glasgow, Scotland"}', name='get_current_weather'), type='function')])

通过不同的提示词，我们可以让它反问不同的问题以获取函数参数信息。

messages = []
messages.append({"role": "system", "content": "Don't make assumptions about what values to plug into functions. Ask for clarification if a user request is ambiguous."})
messages.append({"role": "user", "content": "can you tell me, what is the weather going to be like in Glasgow, Scotland in next x days"})
chat_response = chat_completion_request(
    messages, tools=tools
)
assistant_message = chat_response.choices[0].message
messages.append(assistant_message)
assistant_message

ChatCompletionMessage(content='Sure, I can help with that. Could you please specify how many days ahead you want to know the weather forecast for Glasgow, Scotland?', role='assistant', function_call=None, tool_calls=None)

messages.append({"role": "user", "content": "5 days"})
chat_response = chat_completion_request(
    messages, tools=tools
)
chat_response.choices[0]

Choice(finish_reason='stop', index=0, logprobs=None, message=ChatCompletionMessage(content='', role='assistant', function_call=None, tool_calls=[ChatCompletionMessageToolCall(id='call_b7f3j7im', function=Function(arguments='{"location":"Glasgow, Scotland","num_days":5}', name='get_n_day_weather_forecast'), type='function')]))

并行函数调用

（https://cookbook.openai.com/examples/how_to_call_functions_with_chat_models#parallel-function-calling）

支持一次提问中，并行调用多次函数

messages = []
messages.append({"role": "system", "content": "Don't make assumptions about what values to plug into functions. Ask for clarification if a user request is ambiguous."})
messages.append({"role": "user", "content": "what is the weather going to be like in San Francisco and Glasgow over the next 4 days"})
chat_response = chat_completion_request(
    messages, tools=tools, model=MODEL
)


assistant_message = chat_response.choices[0].message.tool_calls
assistant_message

[ChatCompletionMessageToolCall(id='call_vei89rz3', function=Function(arguments='{"location":"San Francisco, CA","num_days":4}', name='get_n_day_weather_forecast'), type='function'),
ChatCompletionMessageToolCall(id='call_4lgoubee', function=Function(arguments='{"location":"Glasgow, UK","num_days":4}', name='get_n_day_weather_forecast'), type='function')]

使用模型生成函数

（https://cookbook.openai.com/examples/how_to_call_functions_with_chat_models#how-to-call-functions-with-model-generated-arguments）

在这个示例中，演示如何执行输入由模型生成的函数，并使用它来实现可以为我们解答有关数据库的问题的代理。

本文使用Chinook 示例数据库（https://www.sqlitetutorial.net/sqlite-sample-database/）。

指定执行 SQL 查询的函数

（https://cookbook.openai.com/examples/how_to_call_functions_with_chat_models#specifying-a-function-to-execute-sql-queries）

首先，让我们定义一些有用的函数来从 SQLite 数据库中提取数据。

import sqlite3


conn = sqlite3.connect("data/Chinook.db")
print("Opened database successfully")

def get_table_names(conn):
    """Return a list of table names."""
    table_names = []
    tables = conn.execute("SELECT name FROM sqlite_master WHERE type='table';")
    for table in tables.fetchall():
        table_names.append(table[0])
    return table_names




def get_column_names(conn, table_name):
    """Return a list of column names."""
    column_names = []
    columns = conn.execute(f"PRAGMA table_info('{table_name}');").fetchall()
    for col in columns:
        column_names.append(col[1])
    return column_names




def get_database_info(conn):
    """Return a list of dicts containing the table name and columns for each table in the database."""
    table_dicts = []
    for table_name in get_table_names(conn):
        columns_names = get_column_names(conn, table_name)
        table_dicts.append({"table_name": table_name, "column_names": columns_names})
    return table_dicts

现在可以使用这些实用函数来提取数据库模式的表示。

database_schema_dict = get_database_info(conn)
database_schema_string = "\n".join(
    [
        f"Table: {table['table_name']}\nColumns: {', '.join(table['column_names'])}"
        for table in database_schema_dict
    ]
)

与之前一样，我们将为希望 API 为其生成参数的函数定义一个函数规范。请注意，我们正在将数据库模式插入到函数规范中。这对于模型了解这一点很重要。

tools = [
    {
        "type": "function",
        "function": {
            "name": "ask_database",
            "description": "Use this function to answer user questions about music. Input should be a fully formed SQL query.",
            "parameters": {
                "type": "object",
                "properties": {
                    "query": {
                        "type": "string",
                        "description": f"""
                                SQL query extracting info to answer the user's question.
                                SQL should be written using this database schema:
                                {database_schema_string}
                                The query should be returned in plain text, not in JSON.
                                """,
                    }
                },
                "required": ["query"],
            },
        }
    }
]

执行 SQL 查询

（https://cookbook.openai.com/examples/how_to_call_functions_with_chat_models#executing-sql-queries）

现在让我们实现实际执行数据库查询的函数。

def ask_database(conn, query):
    """Function to query SQLite database with a provided SQL query."""
    try:
        results = str(conn.execute(query).fetchall())
    except Exception as e:
        results = f"query failed with error: {e}"
    return results

使用 Chat Completions API 调用函数的步骤：

（https://cookbook.openai.com/examples/how_to_call_functions_with_chat_models#steps-to-invoke-a-function-call-using-chat-completions-api）

步骤 1：向模型提示可能导致模型选择要使用的工具的内容。工具的描述（例如函数名称和签名）在“工具”列表中定义，并在 API 调用中传递给模型。如果选择，函数名称和参数将包含在响应中。

步骤 2：通过编程检查模型是否想要调用函数。如果是，则继续执行步骤 3。
步骤 3：从响应中提取函数名称和参数，使用参数调用该函数。将结果附加到消息中。
步骤 4：使用消息列表调用聊天完成 API 以获取响应。

messages = [{
    "role":"user", 
    "content": "What is the name of the album with the most tracks?"
}]


response = client.chat.completions.create(
    model='myqwen2', 
    messages=messages, 
    tools= tools, 
    tool_choice="auto"
)


# Append the message to messages list
response_message = response.choices[0].message 
messages.append(response_message)


print(response_message)

ChatCompletionMessage(content='', role='assistant', function_call=None, tool_calls=[ChatCompletionMessageToolCall(id='call_23nnhlv6', function=Function(arguments='{"query":"SELECT Album.Title FROM Album JOIN Track ON Album.AlbumId = Track.AlbumId GROUP BY Album.Title ORDER BY COUNT(*) DESC LIMIT 1"}', name='ask_database'), type='function')])

# Step 2: determine if the response from the model includes a tool call.   
tool_calls = response_message.tool_calls
if tool_calls:
    # If true the model will return the name of the tool / function to call and the argument(s)  
    tool_call_id = tool_calls[0].id
    tool_function_name = tool_calls[0].function.name
    tool_query_string = json.loads(tool_calls[0].function.arguments)['query']


    # Step 3: Call the function and retrieve results. Append the results to the messages list.      
    if tool_function_name == 'ask_database':
        results = ask_database(conn, tool_query_string)


        messages.append({
            "role":"tool", 
            "tool_call_id":tool_call_id, 
            "name": tool_function_name, 
            "content":results
        })


        # Step 4: Invoke the chat completions API with the function response appended to the messages list
        # Note that messages with role 'tool' must be a response to a preceding message with 'tool_calls'
        model_response_with_function_call = client.chat.completions.create(
            model="myqwen2",
            messages=messages,
        )  # get a new response from the model where it can see the function response
        print(model_response_with_function_call.choices[0].message.content)
    else: 
        print(f"Error: function {tool_function_name} does not exist")
else: 
    # Model did not identify a function to call, result can be returned to the user 
    print(response_message.content) 

The album "Greatest Hits" contains the most tracks

欢迎点赞关注我，获取更多关于 AI 的前沿资讯。别忘了将今天的内容分享给你的朋友们，让我们一起见证 AI 技术的飞跃！学习商务交流