We are using a python package. Therefore, please run : pip install -e msynth , in the directory that is the parent of the msynth folder.
To run a sweep, run:
wandb sweep sweep.yaml
This will return a sweep id.
wandb agent [your wandb username] / [the project] / [the id given to you by the first command above]
For me thats: wandb agent chiquita/msynth-tests/[id]
# Create a clean conda env
conda create -n msynth python=3.7 -y
conda activate msynth
cd <msynth git repo>
# Install package
pip install -e .
- Install Docker Desktop
- Install X11 for macOS via X Quartz
- Install PulseAudio via homebrew, start the PulseAudio deamon, and verify status:
# Install
brew install pulseaudio
# Start daemon
pulseaudio --load=module-native-protocol-tcp --exit-idle-time=-1 --daemon
# Check pulse audio status
pulseaudio --check -v
- Configure PulseAudio output via:
# List default outputs for pulse audio
pacmd list-sinks
# Set the default output
pacmd set-default-sink <index of sink>
- Try to play an audio file via:
paplay -p <testfile>.wav
- Start Xquartz
open -a Xquartz
- Configure Xquartz to "Allow connections from network clients"
- Whitelist localhost for X11 network connections via paste below on command line
/opt/X11/bin/xhost + "127.0.0.1"
-
Build the docker image into a container docker-compose up --build -d
-
Specify a shared data directory between your local machine and docker to share data
export MSYNTH=<path/for/shared/data>
- Run an existing image
docker-compose up
- Open your IDE of choice
- For Jupyter, open a web browser at http://127.0.0.1:8888 (password is dsp).
- For VS Code, install the Docker extension, then attach to the msyth container and use VS Code.
- For command line SSH access, open a second terminal, find the running container id, and enter it
docker container ls docker exec -it <CONTAINER ID> bash
- Run Linux applications with audio and graphics forwarding.
Enter your container via the SSH access instructions above and launch. Try running reaper within your Linux container on mac via
audacity
reaper