Python - Start Backends

To start the backends, download the YeastMateBackend repository and set up the Python environment as described in Python - Installation. Then activate your conda environment or enter your Docker container.

Detection backend

The detection server requires a model snapshot and a model configuration file. The default configuration files are located in the models directory of the repository. You need to download our model weights file from OSF and add them to the models directory, or alternatively use your own model snapshot.

You can then start the Detection backend with:

python yeastmate_server.py

It comes with multiple optional command line arguments:

  • --gpu : Add this flag if you want to run the Detection on GPU. Default is False.
  • --gpu_index : The index of the GPU to use. Only used if --gpu flag is given. Default is 0.
  • --port : The port on which the backend should listen on. Default is 11005.
  • --config : Path to the config file for the Detection. Only necessary if you re-trained the model. Default is 'models/yeastmate.yaml'.
  • --model : Path to the model weight file for the Detection. Only necessary if you re-trained the model. Default is 'models/yeastmate_weights.pth'.

IO backend

You can start the IO backend with:

python hueyserver.py

It takes the following optional command line argument:

  • --port : The port on which the backend should listen on. Default is 11002.