Run side-by-side prompt engineering and compare as many prompt responses as you need! Try it online at https://promptoctopus.com or self-host.
You'll start off with three "system prompts" you can customize: try filling them with different instructions that test the user. Then add content to the "message" box at the bottom: this is where you'd add the content the user might input. If you're building an Essay Grading AI, you might try three different grading instructions, and submit an essay in the message box.
Hit "Submit" and compare your responses side by side, then try again with more prompts or another message!
Self-Hosting
You'll need an OpenAI API key.
Python
Move /assets into /python_selfhost/static
mv assets python_selfhost/static
Working in /python_selfhost, run Python (ex. with Windows):
pip install virtualenv
virtualenv .venv # or python -m venv .venv
.venv\scripts\activate
# source .venv/bin/activate for mac/unix
Next:
pip install -r requirements.txt
set FLASK_APP=app.py
(use "export" instead of "set" on mac)
create a .env file in /python_selfhost and add:
OPENAI_API_KEY='your-api-key-here'
You're ready to go! Run:
Node.js
In the /nodejs_selfhost folder run:
npm init -y
npm install express cors dotenv openai@^4.0.0
add .env file to project directory and write
OPENAI_API_KEY=your_api_key_here
Move the index.html file from /nodejs_selfhost into your assets folder, then move the folder into /nodejs_selfhost
mv index.html assets
mv assets nodejs_selfhost
node node_script.js
As seen on...
