Project: Technical Evaluation of GPT-4 and GPT-4-Turbo Reflective Listening Response Generation with the Counsel Chat Dataset

This repository contains scripts and data for generating and evaluating chatbot responses using GPT-4 and GPT-4-Turbo models.

Folder Structure

• data/

  • counselchat-data.csv: Original console chat dataset.
  • counsel_chat_data_after_data_preparation.csv: Preprocessed dataset for generating responses.
  • results/
    • v3/generated_responses.csv: Responses generated by GPT-4-Turbo.
    • v4/generated_responses.csv: Responses generated by GPT-4.

• presentation_pics/: Folder for presentation pictures.

• GPT_4_Turbo_model_development.ipynb: Jupyter notebook containing the main script for generating and evaluating responses using GPT-4-Turbo.

• GPT_4_model_development.ipynb: Jupyter notebook containing the main script for generating and evaluating responses using GPT-4.

• requirements.txt: List of required Python packages.

• v1_data_preparation.ipynb: Jupyter notebook for data preprocessing scripts.

Installation

1. Clone the repository:

   git clone git@github.com:dipendrapant/TechnicalEvaluationGPT4.git
   cd TechnicalEvaluationGPT4