Cheaper O1

English

Introduction

Cheaper O1 is a simple Chain-of-Thought (COT) model product designed as an alternative to OpenAI's O1 model. This project utilizes the glm-4-flash model from Zhipu AI to provide a cost-effective solution for complex problem-solving tasks.

Features

• Implements a multi-step reasoning process
• Utilizes web search for information gathering
• Provides self-assessment and judgment of answers

Usage

1. Install the required dependencies:

   pip install -r requirements.txt
   

2. Set up your API key: Create a file named api_key in the project root and paste your Zhipu AI API key.

3. Run the Jupyter notebook:

   jupyter notebook demo.ipynb
   

4. In the notebook, you can use the solve_problem function as shown in the example below:

   from cheaper_o1 import solve_problem
   
   thought_chain = solve_problem("9.11和9.8哪个大")

   This will display the reasoning process and the final answer.

Code Structure

• cheaper_o1.py: Main script containing the core logic
• webviewer.py: Module for web search functionality
• demo.ipynb: Jupyter notebook demonstrating the usage

Note

This project is designed for educational and experimental purposes. It may not be suitable for production environments without further development and testing.

中文

简介

Cheaper O1 是一个简单的思维链（Chain-of-Thought，COT）模型产品，旨在替代 OpenAI 的 O1 模型。本项目使用智谱AI的 glm-4-flash 模型，为复杂问题解决任务提供了一个经济实惠的解决方案。

特性

• 实现多步推理过程
• 利用网络搜索收集信息
• 提供答案的自我评估和判断

使用方法

1. 安装所需依赖：

   pip install -r requirements.txt
   

2. 设置 API 密钥： 在项目根目录创建名为 api_key 的文件，并粘贴您的智谱AI API 密钥。

3. 运行 Jupyter notebook：

   jupyter notebook demo.ipynb
   

4. 在 notebook 中，您可以使用 solve_problem 函数，如下例所示：

   from cheaper_o1 import solve_problem
   
   thought_chain = solve_problem("9.11和9.8哪个大")

   这将显示推理过程和最终答案。

代码结构

• cheaper_o1.py：包含核心逻辑的主脚本
• webviewer.py：网络搜索功能模块
• demo.ipynb：演示使用方法的 Jupyter 笔记本

注意

本项目设计用于教育和实验目的。如果没有进一步的开发和测试，可能不适合用于生产环境.