Custom LLM ChatDoc System

Lunch & Learn: Deploying a Chat LLM System Using Python

Overview This learning guide will walk your colleagues—who have no prior experience in Python, AI, LLMs, or prompt engineering—through the step-by-step process of deploying a Chat LLM system. This is a hands-on experience designed to be playful and accessible, with every detail included for both macOS and Windows environments. The goal is for everyone to have a fully functional LLM system running on their computer by the end of the session.

---

Training Outline

1. Setting up the Environment
   • Python Installation (macOS & Windows)
   • Virtual Environment Setup
   • Installing Dependencies
   • Getting the OpenAI API Key
2. Step-by-Step Guide to Building the Chat LLM App
   • Application Overview
   • Code Walkthrough (app.py, templates, etc.)
   • Running the Application
3. Prompts for Building Each Script
4. Hands-on Practice

Step 1: Environment Setup

1. Install Python

• macOS:

  1. Open Terminal (Cmd + Space, type "Terminal").
  2. Run the command to install Python:

     brew install python3

     Prompt: "How do I install Python on macOS using Terminal?"
  3. Verify installation:

     python3 --version

     Prompt: "How do I check if Python is installed on my Mac?"

• Windows:

  1. Download Python from python.org.
  2. Run the installer, and make sure to check "Add Python to PATH". Prompt: "What are the steps to install Python on Windows and add it to PATH?"
  3. Verify installation by opening Command Prompt (Win + R, type "cmd"):

     python --version

     Prompt: "How can I verify that Python is installed on Windows?"

2. Create and Activate a Virtual Environment

• macOS:

  1. Navigate to your working directory:

     cd ~/Desktop

  2. Create a virtual environment:

     python3 -m venv venv

     Prompt: "How do I create a virtual environment in Python on macOS?"
  3. Activate it:

     source venv/bin/activate

     Prompt: "How do I activate a Python virtual environment on macOS?"

• Windows:

  1. Navigate to your working directory:

     cd Desktop

  2. Create a virtual environment:

     python -m venv venv

     Prompt: "How do I create a virtual environment in Python on Windows?"
  3. Activate it:

     venv\Scripts\activate

     Prompt: "How do I activate a Python virtual environment on Windows?"

3. Install Dependencies

• Both macOS and Windows:
  1. After activating the virtual environment, install the dependencies using the requirements.txt file:

     pip install -r requirements.txt

     Prompt: "How do I install all dependencies from a requirements.txt file?"
  2. If you don’t have requirements.txt, install the required libraries manually:

     pip install openai flask pandas matplotlib PyPDF2 python-dotenv

     Prompt: "What is the command to install Flask, Pandas, Matplotlib, PyPDF2, and python-dotenv using pip?"

4. Obtain OpenAI API Key

• Go to OpenAI and generate your API key.
• Create a .env file in your project directory with the following format to store the API key securely:

  OPENAI_API_KEY=your_openai_api_key_here
  

  Prompt: "How do I create a .env file to securely store my OpenAI API key in a Python project?"

---

Step 2: Building the Chat LLM Application

1. Understand the Application Structure

• app.py: The main Python script that runs the Flask application.
• templates/: HTML files that control how the application pages are rendered.
• static/: Stores static assets like images, JavaScript, and CSS.
• documents/: Stores uploaded documents for processing.

2. Setting Up the Application

• Copy all extracted files (app.py, templates, static, etc.) into your project folder on your desktop. Ensure the .env file is also in this directory.
• Prompt for Setup: "How do I organize a Python Flask project with templates and static files?"

3. Running the Application

• macOS & Windows:
  1. Make sure your virtual environment is activated.
  2. Run the application:

     python app.py

     Prompt: "How do I run a Flask application using Python?"
  3. Open a browser and navigate to http://127.0.0.1:5001 to interact with the chat LLM. Prompt: "Where can I access my Flask application once it is running?"

4. Example Prompts for Building the Application

• To understand each component of the app, ask:

  What does the `app.py` script do in my Flask application? Explain each function and route.
  

• To understand how document uploads are handled:

  Explain the purpose of the `/upload` route in Flask.
  

• To know how the OpenAI API is integrated:

  How does the `/chat` route interact with the OpenAI API?
  

5. Code Walkthrough

• Imports: Explain why each import is necessary (e.g., Flask for the web framework, OpenAI for interacting with the API).
• Routes: Walk through each route in app.py.
  • /upload: Handles document uploads and stores them.
  • /chat: Sends the context to the OpenAI API and displays responses.
  • Prompt for Context: "How does Flask handle file uploads securely?"
• Functions: Discuss helper functions like extract_text() to extract content from uploaded documents.
  • Prompt for Extraction: "How can I extract text from different file types in Python?"

---

Step 3: Hands-on Practice

Exercise 1: Upload and Chat

• Upload a sample document (e.g., a PDF with some text) and ask a question.
• Prompt for Upload: "How do I upload a file and interact with it using a chat feature in Flask?"

Exercise 2: Modify the HTML

• Add a personalized message or title to index.html.
  • Prompt: "How do I edit an HTML file to add a custom title in a Flask app?"

Exercise 3: Add a New Route

• Add a new route called /about that returns information about the application.
  • Prompt: "How do I add a new route in Flask to display an 'About' page?"

---

Deployment Practice

1. Running on a Different Port

• Modify app.py to run on a different port, like 8080:

  if __name__ == '__main__':
      app.run(debug=True, host='127.0.0.1', port=8080)

  Prompt: "How do I change the port of my Flask application?"

2. Creating a Batch File (Windows) and Shell Script (macOS) to Run the App

• Windows:

  1. Open Notepad and paste the following script:

     @echo off
     call venv\Scripts\activate
     python app.py
     pause

  2. Save it as run_app.bat on your desktop.
  3. Double-click the .bat file to run the application.
  • Prompt: "How do I create a batch file to automate running my Python app on Windows?"

• macOS:

  1. Open Terminal and create a new shell script:

     touch run_app.sh

  2. Edit the script:

     nano run_app.sh

  3. Add the following lines:

     #!/bin/bash
     source venv/bin/activate
     python app.py

  4. Save (Ctrl + O) and exit (Ctrl + X).
  5. Make the script executable:

     chmod +x run_app.sh

  6. Run it:

     ./run_app.sh

  • Prompt: "How do I create and run a shell script to automate my Python app on macOS?"

---

Summary and Q&A

• Recap: From installing Python and dependencies to running the app and interacting with an LLM.
• Emphasize that mistakes are okay—every error is a learning opportunity.
• Open the floor for questions, providing playful analogies if needed (e.g., "The virtual environment is like putting on your lab coat before experimenting—keeps things clean!").

Follow-up Resources

• Python Basics: Learn Python
• Flask Documentation: Flask Docs
• OpenAI API Guide: OpenAI API Documentation

Next Steps

• Encourage colleagues to extend the app—maybe add a feature to analyze uploaded text sentiment.
• Plan a follow-up session in a month to review progress and celebrate everyone's learning achievements!