Textless (ASR-transcript free) Spoken Question Answering. The official release of NMSQA dataset and the implementation of "DUAL: Textless Spoken Question Answering with Speech Discrete Unit Adaptive Learning" paper.
Creative Commons Attribution Share Alike 4.0 International