Common repository for The Sound of AI Open Source Research Project.
This main app provides scripts to install and run the end-to-end pipeline. This includes Speech-to-Text, Text-to-Sound, and Sound Generation services, as well as the sampler/playback tool.
A manual that describes the UI functionalities and how to use the app is available at https://github.com/TheSoundOfAIOSR/rg_production/blob/main/Manual.md
If you encounter any issues running or installing the app, please check if there is a Github issue already open for it or open a new issue and get in touch with us for troubleshooting tips.
- Csound https://csound.com/
- latest version should run, tested version in this project is
6.16.2 - https://github.com/csound/csound/releases/tag/6.16.2
- install .exe
- or unzip .zip version in a folder that is added to the system path
On MacOs, if homebrew is installed, the setup script will install Csound automatically.
- Portaudio http://portaudio.com/
On Windows the setup script will install Portaudio. On MacOs, if homebrew is installed, the setup script will install Portaudio automatically.
- Python version 3.8 is to be used
- in case python 3.8 available on the system but not the default one then use
setup3.8.batinstead ofsetup.bat
Each service requires its own virtual python environment which is created during the setup.
C:\Users\User\AppData\Local\Temp\
git clone --recurse-submodules https://github.com/TheSoundOfAIOSR/project_common.git
cd project_common
setup.bat
run.bat
git clone --recurse-submodules https://github.com/TheSoundOfAIOSR/project_common.git
cd project_common
bash setup_mac.sh
bash run_mac.sh
If python3.8 is linked as python3.8 rather than python3, please run setup_mac38.sh instead.
git clone --recurse-submodules https://github.com/TheSoundOfAIOSR/project_common.git
cd project_common
bash setup_linux.sh
bash run.sh
if already cloned without --recurse-submodule
cd project_common
git submodule update --init --recursivefurther updates
git pull --recurse-submodulesThe Sound generator module can also be used via a web app - follow these instructions to get it running.