Whisper models for automatic speech recognition (ASR) and speech translation, quantized for faster inference speeds.