Computational Creativity Submission: Readme and Important Notes

Welcome! This file includes everything you need to know to get the code running. There are two main components:

• Python backend (including the generative model)
• Node.JS server (serving the frontend and communicating with backend)

They interact with each other as indicated here:

The first and only thing to know is that it is: Always start the Node.JS server before the python backend!

Prerequisites

There are a few prerequisites to running the system. The backend requires Python >=3.6 and a virtual environment (https://docs.python.org/3/tutorial/venv.html).

For the frontend Node.JS (https://nodejs.org/en/) is required. The system was tested with version 14, but any version >= 14 should work.

Setup

Once you have Python and Node.JS installed the steps are as follows: These steps assume that the main submission folder is called 'submission'. Node.JS server:

• Open a terminal
• Navigate to the main submission folder
• Execute cd CC_Project_Frontend
• Execute npm i to install the required Node.JS packages
• Execute npm run start to start the Node.JS server
• Proceed with starting the python backend

Python backend: Step 1: Get the pre-trained model

• Download the pre-trained model from https://drive.google.com/file/d/1oX0ggfZiGRePpye6M-rwdPt2msAuei0M/view?usp=sharing
• Place the model.pt file in the submission/PopMusicTransformerPytorch/src/transformer/REMI-my-checkpoint folder.

Step 2: Create python environment

• Create and activate a new virtual python environment (see https://docs.python.org/3/tutorial/venv.html#creating-virtual-environments)
• Open a terminal
• Navigate to the main submission folder
• Execute cd PopMusicTransformerPytorch
• If your python version is 3.6.x: Execute pip install -r requirements_python_3_6.txt
• If your python version is >3.6: Execute requirements_python_3_7+.txt
• Execute cd src/transformer
• Execute python main.py or python3 main.py, depending on your bash aliases

Once both the Node.JS server and the Python backend are running, open a web-browser and go to http://localhost:5000/. Click Generate to generate some MIDI from scratch using the pre-trained model!

Frontend controls

1. Click "Generate" to generate some MIDI from scratch - please be patient with the MIDI generation, it can take a few seconds depending on your CPU/GPU
2. Hit 1, 2 or 3 to select and play track 1, 2 or 3
3. Clicking a track selects it, but doesn't play it back
4. After selecting a track it will be highlighted with a large blurry green border! Hit ENTER to generate new MIDI based on the selected track
5. Go back to step 2 (or 1 if you wanna start from scratch)
6. Repeat until happy
7. The outputs are saved in CC_Project_Frontend/midi after each generation.