当前位置：首页 > news >正文

CLI-Chatbot实现多轮对话以及history

news 2026/6/2 11:34:30

这个项目会让你一次性学会：

LLM API 调用
多轮对话
conversation history
streaming
CLI 工程结构
async 编程
prompt 管理

你最终要做出的结果

类似于：

并且：

一、项目结构

1.先安装uv

uv Documentation

2.创建项目

在cmd终端商执行

uv init cli-chatbot
cd cli-chatbot

3.安装依赖

uv add openai rich python-dotenv

由于LLM本身不记忆历史，每次问问题的时候就会出现

就会非常的麻烦，所以需要加入memory来帮助LLM记忆历史

二、实现多轮对话

memory.py

class ConversationMemory:
def __init__(self):
self.messages = [
{
"role": "system",
"content": "You are a helpful AI assistant."
}
]

def add_user_message(self, content):
self.messages.append({
"role": "user",
"content": content
})

def add_assistant_message(self, content):
self.messages.append({
"role": "assistant",
"content": content
})

def get_messages(self):
return self.messages

app.py

from openai import OpenAI
from dotenv import load_dotenv
from chatbot.memory import ConversationMemory
import os

load_dotenv()

client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))
memory = ConversationMemory()

while True:
user_input = input("\nYou: ")

if user_input.lower() == "exit":
break

memory.add_user_message(user_input)

response = client.chat.completions.create(
model="gpt-4.1-mini",
messages=memory.get_messages()
)

assistant_reply = response.choices[0].message.content

memory.add_assistant_message(assistant_reply)

print("\nAssistant:")
print(assistant_reply)

目前已经实现多轮对话，因为memory.get_messsage()会携带历史记录

stream.py

def stream_response(stream):
full_response = ""

for chunk in stream:
content = chunk.choices[0].delta.content

if content:
print(content, end="", flush=True)
full_response += content

print()

return full_response

修改app.py

from openai import OpenAI
from dotenv import load_dotenv
from chatbot.memory import ConversationMemory
from chatbot.stream import stream_response
import os

load_dotenv()

client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))
memory = ConversationMemory()

while True:
user_input = input("\nYou: ")

if user_input.lower() == "exit":
break

memory.add_user_message(user_input)

stream = client.chat.completions.create(
model="gpt-4.1-mini",
messages=memory.get_messages(),
stream=True
)

print("\nAssistant:\n")

assistant_reply = stream_response(stream)

memory.add_assistant_message(assistant_reply)