Feb 3, 2022 · Our speech translation model demonstrates zero-shot text translation without seeing any text translation data, providing evidence for cross- ...

Feb 3, 2022 · We present mSLAM, a multilingual Speech and. LAnguage Model that learns cross-lingual cross- modal representations of speech and text by pre ...

mSLAM is the multilingual version of SLAM which has been pre-trained on speech data from 51 51 languages and text data from 101 101 languages. mSLAM was ...

We present mSLAM, a multilingual Speech and LAnguage Model that learns cross-lingual cross-modal representations of speech and text by pre-training jointly on ...

A Speech2S model is proposed, which is jointly pre-trained with unpaired speech and bilingual text data for direct speech-to-speech translation tasks, ...

Feb 8, 2022 · New paper: "mSLAM: Massively Multilingual Joint Pre-training for Speech and Text" mSLAM is our new 2B-param speech-text model in 100+ languages ...

Jul 23, 2023 · We present Mu2SLAM, a multilingual sequence-to-sequence model pre-trained jointly on unlabeled speech, unlabeled text and supervised data ...

Oct 31, 2022 · mSLAM: Massively multilingual joint pre-training for speech and text. We present mSLAM, a multilingual Speech and LAnguage Model that learns c..

Joint Pre-Training with Speech and Bilingual Text for Direct Speech to Speech Translation ... mslam: Massively multilingual joint pre-training for speech and text.

Feb 8, 2022 · New paper: "mSLAM: Massively Multilingual Joint Pre-training for Speech and Text" mSLAM is our new 2B-param speech-text model in 100+ languages ...