The goal of the Banana CLI is to bring an npm-like experience to ML development loops.

This version includes:

• banana init to create a project with boilerplate
• banana install to install packages from the requirements.txt
• banana --version to show the current version
• banana deploy to easily deploy the local folder to Banana.dev without using GitHub

Here's a demo video

python3 -m venv .venv
source .venv/bin/activate
pip3 install -e ./
banana --version

1. Create a new project directory with

banana init my-app
cd my-app

2. Start the dev server

python3 app.py

3. Call your API (from a separate terminal)

curl -X POST -H "Content-Type: application/json" -d '{"prompt": "Hello I am a [MASK] model."}' http://localhost:8000/

The interactive dev server works like a react, next, or nodemon server: it selectively hot reloads components when you save changes to different parts of your app.py file.

The init() function is ran on startup and for every change to init().

The handler() function is ran on every change to handler(), without needing to wait for a long init()

4. Try changing the handler, see what happens!
5. Try changing the init, see what happens!

• Lock in a stable interface
• Add the following commands
  • banana stage -> run a temporary deployment to Banana's cluster for on-GPU testing
  • banana test -> unit test against local test cases