chatgpt-comment-analyzer

Use LangChain and Azure OpenAI to analyze social media comments

Introduction

This project is a quick demo for using LangChain and Azure OpenAI to build an LLM application to analyze social media comments. The logic is not complicated, but it is a good example to show how to use prompt template and output parser to effectivly manage the prompt and output, without manipulating the raw prompt string directly.

Installation

1. Clone the repository:

git clone https://github.com/linjungz/chatgpt-comment-analyzer.git
cd chatgpt-comment-analyzer

2. Create python virtual environment:

python3 -m venv .venv
source .venv/bin/activate

3. Install depenancies:

pip install -r requirements.txt

Configuration

1. Obtain your Azure OpenAI API key, Endpoint and Deployment Name from the Azure Portal.

2. Create .env in the root dir and set the environment variable in the file:

OPENAI_API_BASE=https://your-endpoint.openai.azure.com
OPENAI_API_KEY=your-key-here
OPENAI_DEPLOYMENT_NAME=your-gpt-deployment-name

Here's where you can find the deployment names for GPT-3.5 or GPT-4:

Usage

1. The entry for the application is app.py :

$ python app.py --help
usage: app.py [-h] --input INPUT [--sheet SHEET] --output OUTPUT [--count COUNT]

Analyze comments from excel file

optional arguments:
  -h, --help       show this help message and exit
  --input INPUT    input excel file path
  --sheet SHEET    sheet name in the input excel file
  --output OUTPUT  output excel file path
  --count COUNT    number of comments to process

2. The prompt template is in prompt_templates.py which could be adjusted to fit your specific requirement.

3. Some comments extracted from social media are in sample_input.xlsx for testing purpose.