Can I train one streaming speech recognition model that works for multiple languages and lets me choose the language during inference? I want it to be like Whisper, where you can pass the language as a parameter when transcribing audio. The goal is to avoid training separate models for each language and instead have a single model that can handle all of them in real time.