Save the loss information in the file saved with model.save('model.h5')

When training and saving to disk:

history = model.fit(X, Y, batch_size=32, epochs=8, verbose=1)
model.save('model.h5')

and then a few days later, loading it with:

model = load_model('model.h5')

how to find the loss information obtained during the training process?

I have seen in How to return history of validation loss in Keras solutions like saving loss from history.history['loss'] in a separate TXT or JSON file, but instead of keeping a pair of files model.h5 + model.json, how to save this loss information directly in the h5 file? and how to recover it later?