「Ollama · part 2」LangChain-Ollama · 步骤及演示-CSDN博客

本文链接：https://blog.csdn.net/OvO_ll/article/details/148849235

部署大模型方法大同小异

创建文件

requirements.txt

langchain-core
langchain-community
langchain-ollama
python-dotenv

.env

OLLAMA_MODEL=llama3

main.py

from langchain_community.llms import Ollama
from langchain_core.prompts import ChatPromptTemplate 
from langchain_core.output_parsers import StrOutputParser 
import requests  # 用于测试连接

def check_ollama_connection():
    """检查Ollama服务是否运行"""
    try:
        response = requests.get("http://localhost:11434")
        if response.status_code == 200:
            print("Ollama服务已启动")
            return True
    except requests.ConnectionError:
        print("无法连接到Ollama服务")
        print("请确保Ollama正在运行: `ollama serve`")
        return False

def check_model_available(model_name):
    """检查模型是否可用"""
    try:
        response = requests.post(
            "http://localhost:11434/api/show",
            json={"name": model_name}
        )
        if response.status_code == 200:
            print(f"模型 '{model_name}' 已加载")
            return True
    except Exception:
        pass
    
    print(f"模型 '{model_name}' 未找到")
    print(f"请下载模型: `ollama pull {model_name}`")
    return False

# 配置模型 - 可以选择较小的模型如 'phi3' 如果llama3太大
MODEL_NAME = "llama3"

# 检查服务状态
if not check_ollama_connection() or not check_model_available(MODEL_NAME):
    exit(1)

# 初始化Ollama
llm = Ollama(
    model=MODEL_NAME,
    base_url="http://localhost:11434",
    temperature=0.7,
    system="你是一个耐心的编程导师，用简单的中文解释概念"
)

# 创建对话链
prompt = ChatPromptTemplate.from_template(
    "用简单的中文回答这个编程问题：{input}"
)
chain = prompt | llm | StrOutputParser()

# 运行对话
print("\n示例问答：")
response = chain.invoke({"input": "编程学习有哪些要求"})
print("\nAI回答：", response)

# 流式输出演示
print("\n流式输出演示（解释神经网络）：")
for chunk in chain.stream({"input": "用简单的中文解释神经网络"}):
    print(chunk, end="", flush=True)

终端运行

python main.py