Add support for the Voxtral-Mini-4B-Realtime-2602 model from Hugging Face via the vLLM backend.
Model URL: https://huggingface.co/mistralai/Voxtral-Mini-4B-Realtime-2602
This model is designed for real-time automatic speech recognition (ASR) and benefits from streaming inference capabilities. Integration with LocalAI's vLLM backend would enable real-time transcription workflows. This may require extending the current gRPC streaming interface to properly support real-time ASR streaming.