Text this: Efficient Prediction of Shallow-Water Acoustic Transmission Loss Using a Hybrid Variational Autoencoder–Flow Framework