OpenAI Chatbot using Python

A Python-based conversational AI chat application that seamlessly integrates with OpenAI models or compatible local LLM services. This project allows you to have interactive conversations while maintaining full conversation history.

---

Table of Contents

• Features
• Prerequisites
• Installation
• Configuration
• Usage
• Project Structure
• Advanced Configuration
• Troubleshooting
• Contributing
• License

---

Features

✨ Core Features:

• 💬 Interactive conversational CLI interface with persistent message history
• 🔌 Flexible LLM backend support (OpenAI, local models, or compatible endpoints)
• ⚙️ Highly configurable model parameters (temperature, max tokens, etc.)
• 🛡️ Secure API key management with environment variable support
• 📝 Conversation context preservation for multi-turn interactions
• 🚀 Simple and intuitive command-line interface

---

Prerequisites

Before you begin, ensure you have the following installed on your system:

• Python: Version 3.8 or higher
• pip: Python package manager (comes with Python)
• Git: For cloning the repository (optional)

Optional Requirements

Depending on your LLM backend choice:

• Local LLM Service: If using local models, ensure a compatible LLM service is running (e.g., LM Studio, Ollama, or similar)
• OpenAI API Key: If using OpenAI's API directly instead of a local service

---

Installation

1. Clone the Repository

git clone https://github.com/sudipta-chaudhari/ai-chat.git
cd ai-chat

2. Create a Virtual Environment (Recommended)

On Windows:

python -m venv venv
venv\Scripts\activate

On macOS/Linux:

python -m venv venv
source venv/bin/activate

3. Install Dependencies

pip install -e .

Or install dependencies manually:

pip install "openai>=1.3.0"

---

Configuration

The project uses a Settings class to manage all LLM configuration parameters. Configuration is defined in src/settings.py.

Required Parameters

The Settings class requires all 5 parameters to be provided during initialization (no defaults):

Parameter	Description	Type	Constraints
base_url	API endpoint URL	str	Must start with http or https
api_key	Authentication key	str	Any string (local models may use "not needed")
model	Model identifier	str	Model name/ID
temperature	Response creativity	float	0.0 - 1.0
max_tokens	Max response tokens	int	Positive integer

Customizing Configuration

Configure settings in main.py when creating the Settings instance:

settings = Settings(
    base_url="http://127.0.0.1:1234/v1",
    api_key="not needed",
    model="liquid/lfm2.5-1.2b",
    temperature=0.7,
    max_tokens=512
)

All configuration values can be modified after creation using property setters:

settings.temperature = 0.9  # Adjust creativity
settings.max_tokens = 1024  # Increase response length

Using Environment Variables (Recommended for Production)

For better security, especially for API keys, use environment variables in main.py:

import os

settings = Settings(
    base_url=os.getenv("LLM_BASE_URL", "http://127.0.0.1:1234/v1"),
    api_key=os.getenv("LLM_API_KEY", "not needed"),
    model=os.getenv("LLM_MODEL", "liquid/lfm2.5-1.2b"),
    temperature=float(os.getenv("LLM_TEMPERATURE", "0.7")),
    max_tokens=int(os.getenv("LLM_MAX_TOKENS", "512"))
)

Then set environment variables:

On Windows (PowerShell):

$env:LLM_API_KEY = "your-api-key-here"
$env:LLM_BASE_URL = "https://api.openai.com/v1"
$env:LLM_MODEL = "gpt-3.5-turbo"

On macOS/Linux (Bash):

export LLM_API_KEY="your-api-key-here"
export LLM_BASE_URL="https://api.openai.com/v1"
export LLM_MODEL="gpt-3.5-turbo"

---

Usage

Running the Application

python main.py

Example Interaction

🤖 Chat initialized with model: liquid/lfm2.5-1.2b
Type 'exit' to quit, 'clear' to clear conversation history
--------------------------------------------------

You: What is Machine Learning? Explain in one sentence. 
Assistant: Machine learning is a subset of artificial intelligence that enables computers to learn patterns from data without being explicitly programmed.

You: Write one more sentence.
Assistant: Machine learning empowers systems to improve their performance on tasks through experience and data analysis.

You: exit
Goodbye!

Commands

Command	Action
Type any question	Send message to the chat
clear	Clear conversation history
exit	Quit the application

---

Project Structure

ai-chat/
├── main.py                          # Application entry point
├── pyproject.toml                   # Project metadata and dependencies
├── README.md                        # This file is primary documentation for a project
├── src/
│   ├── __init__.py                 # Package initialization
│   ├── settings.py                 # Configuration settings (Settings class)
│   └── chat/
│       ├── __init__.py             # Package initialization
│       ├── chat_client.py          # OpenAI client wrapper
│       └── chat_session.py         # Conversation history management
└── openai_chat.egg-info/           # Package metadata (generated)

File Descriptions

• main.py: Implements the CLI interface and main application loop
• src/settings.py: Settings class with LLM configuration parameters
• src/chat/chat_client.py: ChatClient class that handles API communication
• src/chat/chat_session.py: ChatSession class for managing conversation history
• pyproject.toml: Project metadata, version, and dependencies

---

Advanced Configuration

Setting Up with OpenAI's API

To use OpenAI's official API instead of a local service:

1. Create an OpenAI account at https://openai.com
2. Generate an API key from your account dashboard
3. Configure settings in main.py when creating the Settings instance:

settings = Settings(
    base_url="https://api.openai.com/v1",
    api_key="your-api-key-here",  # IMPORTANT: For production, use environment variables as shown above!
    model="gpt-3.5-turbo",  # or "gpt-4"
    temperature=0.7,
    max_tokens=512
)

Setting Up with Local LLM Services

Using LM Studio

1. Download from https://lmstudio.ai
2. Download a model in the LM Studio interface
3. Start the local server (usually runs on http://127.0.0.1:1234)
4. Keep default configuration in src/settings.py (or override in main.py).

Using Ollama

1. Install from https://ollama.ai
2. Pull a model: ollama pull llama2
3. Models run on http://127.0.0.1:11434 by default
4. Update src/settings.py (or override in main.py):

# Example of overriding in main.py, before initializing ChatClient:
settings.base_url = "http://127.0.0.1:11434/v1"
settings.model = "llama2"

Optimizing for Different Use Cases

You can optimize the model's behavior by adjusting temperature and max_tokens settings. These are defined as default values in src/settings.py and can be overridden in main.py before initializing the ChatClient.

For Factual/Precise Responses:

# Default in src/settings.py or override in main.py before initializing ChatClient:
settings.temperature = 0.1  # Lower temperature = more deterministic/focused
settings.max_tokens = 256   # Shorter responses for conciseness

For Creative Responses:

# Default in src/settings.py or override in main.py before initializing ChatClient:
settings.temperature = 0.9  # Higher temperature = more creative/varied
settings.max_tokens = 1024  # Longer responses for richer content

For Balanced Performance (Default):

# Default settings in src/settings.py:
settings.temperature = 0.7  # Moderate creativity and consistency
settings.max_tokens = 512   # Balanced response length

---

Troubleshooting

Issue: "Connection refused" error

Cause: LLM service is not running or URL is incorrect

Solution:

1. Verify the LLM service is running on the configured URL
2. Check base_url in src/settings.py
3. Test the endpoint with: curl http://127.0.0.1:1234/v1/models

Issue: "Authentication failed" error

Cause: Invalid or missing API key

Solution:

1. For local models: Set api_key = "not needed" in src/settings.py
2. For OpenAI: Verify your API key is correct and active
3. Check for typos or extra whitespace in the key

Issue: Slow responses or timeouts

Cause: Model is processing-intensive or network issues

Solution:

1. Reduce max_tokens in src/settings.py
2. Lower temperature for simpler processing
3. Check network connectivity
4. Try a smaller or faster model

Issue: ImportError for 'openai' module

Cause: Dependencies not installed

Solution:

pip install -e .
# or
pip install "openai>=1.3.0"

Issue: Model not found error

Cause: Specified model is not available

Solution:

1. Verify the model name in model
2. Ensure the model is downloaded/installed on your system
3. Check available models: ollama list or your service's model manager

---

Contributing

Contributions are welcome! Please follow these guidelines:

1. Fork the repository on GitHub
2. Create a feature branch: git checkout -b feature/your-feature-name
3. Make your changes with clear, descriptive commits
4. Test thoroughly before submitting
5. Push to your fork: git push origin feature/your-feature-name
6. Submit a Pull Request with a description of your changes

Code Style

• Follow PEP 8 guidelines
• Use meaningful variable and function names
• Add docstrings to functions
• Keep functions focused and single-purpose

---

License

This project is open-source and available under the MIT License. See LICENSE file for details.

---

Support & Contact

• 📧 Email: Not provided
• 🐛 Issues: Report bugs on the GitHub Issues page
• 💬 Discussions: Join conversations in GitHub Discussions

---

Changelog

Version 0.1.0 (Current)

• Initial release
• Basic CLI interface
• OpenAI API integration
• Local LLM support
• Conversation history management

---

Acknowledgments

• Built with OpenAI Python SDK
• Compatible with local LLM services like LM Studio and Ollama
• Inspired by modern AI chat interfaces

---

Last Updated: March 2026