Tensorflow TimeSeries prediction

Question

I am currently trying to learn timeseries predictions with tensorflow. I have followed the guide in the tensorflow page:

https://www.tensorflow.org/tutorials/structured_data/time_series

I have applied it to my own dataset and everything seems to work, but I have a question that the tutorial doesn't cover and that is the actual prediction. So suppose we have created a data window with:

wide_window = WindowGenerator(input_width=24, 
                              label_width=24,
                              shift=1,label_columns=[(degC)'])

Then, built the model, compile, fit and get the mae:

lstm_model=tf.keras.models.Sequential([
           tf.keras.layers.LSTM(32,return_sequences=True),
           tf.keras.layers.Dense(units=1)])

history = compile_and_fit(lstm_model, wide_window)

val_performance['LSTM'] = lstm_model.evaluate(wide_window.val)
performance['LSTM'] = lstm_model.evaluate(wide_window.test, verbose=0)`

Now, how to use the lstm_model.predict() method, for example, in the test_df, or any other df? I am currently stuck and can't seem to find an answer.