Plotting Bayesian Optimization results based on TensorBoard logs

[Seyedfarazeftehar]

Hi,

I have a question regarding plotting the Bayesian Optimization results with TensorBoard logs.
My codes is based on the below, however when I try to use Tensorboard logs to plot Bayesian Optimization process, it does not show anything.

Could you please guide me how can I plot Bayesian Optimization process with TensorBoard logs?

Code:

def build_model(hp):
    model = Sequential()
    model.add(Input((5,)))  
    model.add(Dense(units=hp.Choice("units1", [256, 512, 1024])))
    model.add(Activation(activations.relu))
    model.add(Dense(units=hp.Choice("units2", [128, 256, 512, 1024])))
    model.add(Activation(activations.relu))
    model.add(Dense(units=hp.Choice("units3", [128, 256, 512, 1024])))
    model.add(Activation(activations.relu))
    model.add(Dense(units=hp.Choice("units4", [128, 256, 512, 1024])))
    model.add(Activation(activations.relu))
    #output layer
    model.add(Dense(1, activation='linear'))
    model.compile(loss= MeanSquaredError(reduction='sum_over_batch_size', name="mean_squared_error"),
                  optimizer = Adamax(learning_rate=0.001) ,
                  metrics=[RootMeanSquaredError(name="root_mean_squared_error")])
                  
    return model

build_model(keras_tuner.HyperParameters())

tuner = keras_tuner.BayesianOptimization(
    hypermodel=build_model,
    objective=keras_tuner.Objective("val_loss", direction="min"),
    max_trials=3000,
    alpha=0.0001,
    beta=500,
    seed =0,
    overwrite=True,
    directory="learning_rate",
    project_name="learning_rate",
)


stop_early = EarlyStopping(monitor='val_loss', patience=5)

tuner.search(x_scaled, y, epochs=500, batch_size= 128, validation_split=0.2, shuffle= True, callbacks=[stop_early, TensorBoard("logs")])

tuner.results_summary()