Keras Tuner: Accessing the data in the directory created during Hyperparameter Tuning using the Keras Tuner in Google Colab

Question

I am tuning an ANN model using the Keras Tuner, using the Bayesian optimizer as follows:

tuner = BayesianOptimization (
    build_model,
    objective=kt.Objective('val_auc', direction='max'),
    max_trials=30,
    executions_per_trial=10,
    directory='my_dir',
    project_name='silent_sound_ann_bayesian_search_1.5')

The code runs, but I want to access the data stored for each of the 10 executions per trial. I can access the checkpoints and trial.json files for each trial via the Colab file structure in directory='my_dir', but it doesn't seem like the AUC score for each of the 10 executions per trial is logged.

There are 30 trial.json files in directory='my_dir', one for each hyperparameter configuration tested. But each of those configurations was executed 10 times, which means there should be 10 AUC values stored in each trial.json file.

I want to plot box plots of the AUC scores of the executions for each hyperparameter configuration, and doing so requires access to the scores of each execution. Any suggestions on where I might find that data? It doesn't seem to be stored in the oracle.json or the tuner.json files either. Or a more efficient way of capturing the execution scores as the model runs each trial?