3 minOpen Source vs. Usecase-Specific Models: Why Tailored Solutions in Speech Emotion Recognition (SER)