https://github.com/helgesverre/python-music-generation
Various scripts to generate music
https://github.com/helgesverre/python-music-generation
Last synced: 21 days ago
JSON representation
Various scripts to generate music
- Host: GitHub
- URL: https://github.com/helgesverre/python-music-generation
- Owner: HelgeSverre
- Created: 2024-09-27T04:28:25.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2024-10-02T08:14:25.000Z (about 1 year ago)
- Last Synced: 2025-01-09T02:51:52.326Z (9 months ago)
- Language: Python
- Homepage:
- Size: 3.69 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Python Scripts for Music Generation
[](https://github.com/HelgeSverre/python-music-generation)
[](https://www.python.org/downloads/)
[](https://opensource.org/licenses/MIT)
## 🎵 Project Description
This repository contains a diverse collection of Python scripts designed to generate MIDI files for music creation.
All code in this project is **AI-generated**, showcasing various approaches to algorithmic music composition, including
Markov chains, machine learning models, and rule-based systems.
**Note:** This project is intended for experimentation, learning, and exploration of AI-driven music generation
techniques, rather than professional music production.
## 🚀 Getting Started
### Prerequisites
- Python 3.7 or higher
- pip (Python package manager)
- Virtual environment (recommended)
### Installation
1. Clone the repository:
```shell
git clone https://github.com/HelgeSverre/python-music-generation.git
cd python-music-generation
```
2. Create and activate a virtual environment:
```shell
python -m venv venv
source venv/bin/activate # On Windows, use `venv\Scripts\activate`
```
3. Install dependencies (might be outdated):
```shell
pip install -r requirements.txt
```
4. For Mac users, install TensorFlow with Metal support:
```shell
pip install tensorflow-macos==2.13.0
pip install tensorflow-metal==1.0.0
```
## 📁 Project Structure
```
.
├── README.md
├── docs/
├── libs/
├── markov/
├── models/
├── output/
│ ├── images/
│ ├── markov_models/
│ └── midi/
├── random/
├── tensorflow/
└── tools/
```
### ❗ Not included in this repository
- `trance_midis/`: MIDI files used by various scripts, not included in this repository as it contains copyrighted
material. Replace it with your own MIDI file collection.
### Key Directories and Files
- `docs/`: Documentation files, including `transformer_trance.md`
- Explains the Transformer model for trance music generation (AI-generated documentation, use with caution)
- `libs/`: Utility functions and helper scripts
- `utils.py`: Utility functions for music generation (e.g., getting unique file names with number suffixes)
- `markov/`: Markov chain-based music generation scripts
- `models/`: Saved TensorFlow models for music generation
- `output/`: Generated MIDI files, images, and Markov models
- `random/`: Various music generation scripts using different techniques
- `tensorflow/`: TensorFlow-based music generation scripts (currently broken)
- `tools/`: Utility tools, including MIDI to CSV converter
## 🎹 Running the Scripts
1. Ensure your virtual environment is activated.
2. Navigate to the desired script directory (e.g., `random/`, `tensorflow/`, or `markov/`).
3. Run a script:
```shell
python script_name.py
```
For example:
```shell
python random/trance.py
```
4. Check the `output/midi/` directory for generated MIDI files.
## 🧠 Scripts Overview
1. **Markov Chain Models** (`markov/`):
- `markov.py`, `markov_2.py`, `markov_3.py`: Different implementations of Markov chain-based music generation
2. **Random Generation** (`random/`):
- `claude.py`: Claude AI-generated music script
- `trance.py`, `trance_2.py`, ...: Various trance music generation scripts using different rule-based generation
techniques
- `uplifting.py`, `uplifting_2.py`: Uplifting trance generation scripts
- Other scripts for generating hardstyle, arpeggios, and more
3. **TensorFlow Models** (`tensorflow/`): (currently broken)
- `tensorflow_trance_midi.py`: TensorFlow-based trance MIDI generation
- `transformer_trance.py`: Transformer model for trance music generation
4. **Utility Scripts**:
- `tools/midi-to-csv.py`: Convert MIDI files to CSV format (
`midi_note ,note_name ,event ,velocity ,delta ,cumulative_time`)
- `libs/utils.py`: Utility functions used across scripts
## 🎧 Output
Generated MIDI files can be found in the `output/midi/` directory. You can use various software to play these files:
- [VLC Media Player](https://www.videolan.org/vlc/) (free, cross-platform)
- [MuseScore](https://musescore.org/) (free, open-source music notation software)
- Your preferred Digital Audio Workstation (DAW)
## 🛠️ Development
To format the code with the Black code style, run:
```shell
pipx run black *.py
```
## 🤝 Contributing
While this project is primarily an AI experiment, contributions are welcome! If you have ideas for improvements or new
AI-generated scripts, feel free to open an issue or submit a pull request.
## 📄 License
This project is licensed under the MIT License. See the [LICENSE](LICENSE) file for details.
## 🤖 AI Models Used
- **OpenAI**: ChatGPT (`gpt-4o` and `o1-preview`)
- **Anthropic**: Claude Sonnet 3.5 (`claude-3-5-sonnet-20240620`)
---
**Disclaimer:** The code in this repository, including the README, is _mostly_ AI-generated and may contain errors or inconsistencies. Use at your own
discretion and always review the code before running it in your environment.