English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

We present Mu$^{2}$SLAM, a multilingual sequence-to-sequence modelpre-trained jointly on unlabeled speech, unlabeled text and supervised dataspanning Automatic Speech Recognition (ASR), Automatic Speech Translation (AST)and Machine Translation (MT), in over 100 languages. By leveraging a quantizedrepresentation of speech as a target, Mu$^{2}$SLAM trains the speech-textmodels with a sequence-to-sequence masked denoising objective similar to T5 onthe decoder and a masked language modeling (MLM) objective on the encoder, forboth unlabeled speech and text, while utilizing the supervised tasks to improvecross-lingual and cross-modal representation alignment within the model. OnCoVoST AST, Mu$^{2}$SLAM establishes a new state-of-the-art for models trainedon public datasets, improving on xx-en translation over the previous best by1.9 BLEU points and on en-xx translation by 1.1 BLEU points. On Voxpopuli ASR,our model matches the performance of an mSLAM model fine-tuned with an RNN-Tdecoder, despite using a relatively weaker sequence-to-sequence architecture.On text understanding tasks, our model improves by more than 6\% over mSLAM onXNLI, getting closer to the performance of mT5 models of comparable capacity onXNLI and TydiQA, paving the way towards a single model for all speech and textunderstanding tasks.

Mu$^{2}$SLAM: Multitask, Multilingual Speech and Language Models