A Weakly-Supervised Streaming Multilingual Speech Model with Truly Zero-Shot Capability

Open in new window