After having ran an Hyperparameter search with Keras like so:
tuner = kt.BayesianOptimization(
build_model,
objective=kt.Objective("my_obj_fn", direction="min"),
max_trials=5,
directory=files_dir+'my_dir')
X_train, y_train, x_test, y_test = bot.return_data()
tuner.search(X_train, y_train, epochs=5, validation_data=(x_test, y_test))
tuner.results_summary(num_trials=10)
the results/checkpoints are in subdirectories in my_dir named trial_****. The function results_summary() displays for each trials theses lines:
Trial summary
Hyperparameters:
units: 128
rate: 0.0001
learning_rate: 0.01
Score: 0.0036238597240298986
but it DOES NOT specify to which trial_*** directory it belongs to.
Any cue ?
Regards