RT @kwindla: NVIDIA just released a new open source transcription model, Nemotron Speech ASR, designed from the ground up for low-latency u…