A SpeechBrain model fine-tuned from Wav2Vec2 for categorical emotion recognition on the IEMOCAP dataset.