A repository for training machine learning models and using them to find and tune MKID resonators.
train_model.py is a script that takes an mlDict file (yaml format) and creates a machine learning model from it. -The mlDict.yml file is an example model configuration file that should be replaced with user specified settings, data files, and directories.
findResonatorsWPS.py is the main ML inference script; it takes an ML model and 2D (freq x atten) widesweep data (e.g. as generated by freqSweep.py in MKIDReadout), and generates corresponding metadata specifying the list of resonator frequencies and optimal drive powers.
For help with running these scripts and the ML procedure, refer to HELP.md