Based on spectraldoy/music-transformer

A heavily modified version of Music Transformer with a graphical user interface (GUI) added. https://github.com/spectraldoy/music-transformer

What's new?

I've optimized the source code to run on internal hardware, reducing VRAM consumption and increasing GPU acceleration. I've added a graphical interface and one-click installation scripts. I've also revamped the folder management. Simply put, it's now easier to use and install.

Setting up

Clone, the git repository, cd into it if necessary, and install the requirements. Then you're ready to preprocess MIDI files, as well as train and generate music with a Music Transformer.

git clone https://github.com/kilugit/Music-Transformer-GUI
cd Music-Transformer-GUI

Manual Installation:

Create env:

py -3.12 -m venv venv
venv/Scripts/activate

For AMD GPUs (ROCm):

pip install --no-cache-dir ^
    https://repo.radeon.com/rocm/windows/rocm-rel-7.2.1/rocm_sdk_core-7.2.1-py3-none-win_amd64.whl ^
    https://repo.radeon.com/rocm/windows/rocm-rel-7.2.1/rocm_sdk_devel-7.2.1-py3-none-win_amd64.whl ^
    https://repo.radeon.com/rocm/windows/rocm-rel-7.2.1/rocm_sdk_libraries_custom-7.2.1-py3-none-win_amd64.whl ^
    https://repo.radeon.com/rocm/windows/rocm-rel-7.2.1/rocm-7.2.1.tar.gz
	
pip install --no-cache-dir ^
    https://repo.radeon.com/rocm/windows/rocm-rel-7.2.1/torch-2.9.1%2Brocm7.2.1-cp312-cp312-win_amd64.whl ^
    https://repo.radeon.com/rocm/windows/rocm-rel-7.2.1/torchaudio-2.9.1%2Brocm7.2.1-cp312-cp312-win_amd64.whl ^
    https://repo.radeon.com/rocm/windows/rocm-rel-7.2.1/torchvision-0.24.1%2Brocm7.2.1-cp312-cp312-win_amd64.whl
	
pip install -r requirements.txt

For Intel Arc GPUs (XPU):

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/xpu
pip install -r requirements.txt

For AMD/Intel/NVIDIA GPUs (DirectML):

pip install torch-directml
pip install -r requirements.txt

Launch GUI:

run.bat (Windows)

or

venv/Scripts/activate
py main.py