This is a model for Taiwanese audio recognition using LoRA on Whisper-large-v2.