An implemetation of "SmartGPT" as proposed by AI explained in this video: https://www.youtube.com/watch?v=wVzuvf9D9BU&t=434s
✔️ The ability to use different models for each part of the way is available (You can use gpt3.5 and 4)
✔️ Offers the ability to change the # of outputs produced initially for the question
✔️ Automatically saves the messages after the script finishes
Clone the repository:
git clone https://github.com/JarodMica/SmartGPT.git
Navigate into the repository folder and then set up the needed libraries. I recommend that you use a venv to do so and then run:
pip install requirements.txt
Setup your openai key in keys.yaml.example. You will have to change the name to keys.yaml in order for this to work.
As well, you will need to enter your own API key which can be found on openai's website @https://platform.openai.com/overview
Once you have everything set-up, navigate to the folder in CMD and then run the script with:
python smartGPT.py
If you're using an IDE, run it as you would normally (debug). If you're using a venv, make sure you have actually activated it as well!
- Exact implementation may differ than that of what was discussed in Phillip's video.
- Credit to this exact implementation goes to Phillip