SmartGPT - An Efficient Approach to OpenAI's ChatGPT

This repository contains a Python application that interacts with the OpenAI GPT-3.5-turbo model in an efficient manner. It utilizes a multi-step approach to generate, evaluate, and refine responses from the model.

Contents

• config/ - This folder contains the Hydra configuration files for the application.
• SmartGPT.py - This is the main Python script that runs the application.
• openai_api_chat_completion.py - This is a Python module that contains classes and functions for interacting with the OpenAI API.
• environment.yml - This file contains the conda environment configuration.

Installation

1. Clone this repository:

git clone https://github.com/munirjojoverge/SmartGPT.git

2. Navigate into the cloned repository:

cd SmartGPT

3. Create a new conda environment from the environment.yml file:

conda env create -f environment.yml

4. Activate the new environment:

conda activate smartgpt

5. Copy your OpenAI key into the appropriate config file in the config/ directory.

Usage

You can run the application by executing the SmartGPT.py script:

python SmartGPT.py

This will launch the application and begin the interaction with the GPT-3.5-turbo model.

Promp Frameworks:

Chain of Thought

Answer: Let's work this out in a step by step way to be sure we have the right answer

Relexion: Discover flaws and faulty logic

You are a researcher tasked with investigating the X response options provided. List the flaws and faulty logic of each answer option. Let's work this out in a step by step way to be sure we have all the errors:

Resolution

You are a resolver tasked with 1) finding which of the X answer options the researcher thought was best 2) improving that answer, and 3) Printing the improved answer in full. Let's work this out in a step by step way to be sure we have the right answer:

References

• Automatically Discovered Chain of Thought: https://arxiv.org/pdf/2305.02897.pdf
• Karpathy Tweet: https://twitter.com/karpathy/status/1...
• Best prompt: Theory of Mind: https://arxiv.org/ftp/arxiv/papers/23...
• Few Shot Improvements: https://sh-tsang.medium.com/review-gp...
• Dera Dialogue Paper: https://arxiv.org/pdf/2303.17071.pdf
• MMLU: https://arxiv.org/pdf/2009.03300v3.pdf
• GPT 4 Technical report: https://arxiv.org/pdf/2303.08774.pdf
• Reflexion paper: https://arxiv.org/abs/2303.11366
• Why AI is Smart and Stupid: https://www.youtube.com/watch?v=SvBR0OGT5VI&t=1s
• Lennart Heim Video: https://www.youtube.com/watch?v=7EwAdTqGgWM&t=67s

---

This repo was inspired by: @ai-explained

https://youtu.be/wVzuvf9D9BU