CodeLlama Chatbot with Document Context

This Gradio-based chatbot application utilizes the CodeLlama-34b-Instruct-hf model from Hugging Face for conversational interactions. It allows users to upload .txt or .docx files to provide additional context for the chatbot's responses.

Features

• Chat Interface: A user-friendly chat interface powered by Gradio.
• Document Context: Users can upload .txt or .docx files to provide context for the chatbot.
• CodeLlama Model: Uses the CodeLlama-34b-Instruct-hf model for generating responses.
• Parameter Tuning: Users can adjust max_tokens and temperature for response generation.
• Error Handling: Robust error handling for network issues, invalid inputs, and unexpected exceptions.
• Logging: Uses Python's logging module to log execution details and errors.
• Environment Variables: Uses .env files to securely manage the Hugging Face API key.

Prerequisites

• Python 3.12
• pip (Python package installer)
• Hugging Face API key (stored in a .env file)

Installation

1. Clone the repository (if applicable):

   git clone <your-repository-url>
   cd <your-repository-directory>

2. Install dependencies:

   pip install gradio huggingface_hub python-docx python-dotenv

3. Create a .env file:

   • In the same directory as your Python script, create a file named .env.
   • Add your Hugging Face API key to the .env file:

     HUGGINGFACE_API_KEY=your_huggingface_api_key
     

     • Replace your_huggingface_api_key with your actual API key.

4. Run the application:

   python your_script_name.py

   • Replace your_script_name.py with the actual name of your Python script.

Usage

1. Launch the application: Run the Python script.
2. Chat with the chatbot: Enter your messages in the chat interface.
3. Upload a file (optional): Upload a .txt or .docx file to provide additional context.
4. Adjust parameters: Use the sliders to adjust max_tokens and temperature.
5. View responses: The chatbot's responses will appear in the chat interface.

Code Explanation

• respond(current_msg, history_msg, max_tokens, temperature, file):
  • This function handles the chatbot's response generation.
  • It takes the current message, chat history, max_tokens, temperature, and uploaded file as input.
  • It constructs the system message, including any context from the uploaded file.
  • It uses the Hugging Face InferenceClient to interact with the CodeLlama model.
  • It handles potential errors and logs execution details.
• chatbot = gr.ChatInterface(...):
  • This creates the Gradio chat interface.
  • It specifies the respond function as the chatbot's response generator.
  • It adds input components for max_tokens, temperature, and file upload.
• .env and API Key:
  • The .env file and load_dotenv() are used to manage the Hugging Face API key securely.
  • This prevents the API key from being hardcoded in the script.
• File Handling:
  • The code handles .txt and .docx files, extracting the text content for context.
  • The python-docx library is used for .docx files.
• Logging:
  • The logging module logs execution details and errors.

Dependencies

• gradio
• huggingface_hub
• python-docx
• python-dotenv

Notes

• Ensure your Hugging Face API key is correctly set in the .env file.
• The CodeLlama-34b-Instruct-hf model might require significant resources.
• Error handling is included, but further testing is recommended.
• The current time placeholder in the file output needs to be replaced with the actual current time implementation.