The reviewed record of science sign in
Pith

arxiv: 2310.05224 · v1 · pith:UJCCHBZO · submitted 2023-10-08 · cs.CL · cs.LG

Generative Spoken Language Model based on continuous word-sized audio tokens

Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel pith:UJCCHBZOrecord.jsonopen to challenge →

classification cs.CL cs.LG
keywords languagemodelembeddingsgenerativelexicalspokenaudiocontinuous
0
0 comments X
read the original abstract

In NLP, text language models based on words or subwords are known to outperform their character-based counterparts. Yet, in the speech community, the standard input of spoken LMs are 20ms or 40ms-long discrete units (shorter than a phoneme). Taking inspiration from word-based LM, we introduce a Generative Spoken Language Model (GSLM) based on word-size continuous-valued audio embeddings that can generate diverse and expressive language output. This is obtained by replacing lookup table for lexical types with a Lexical Embedding function, the cross entropy loss by a contrastive loss, and multinomial sampling by k-NN sampling. The resulting model is the first generative language model based on word-size continuous embeddings. Its performance is on par with discrete unit GSLMs regarding generation quality as measured by automatic metrics and subjective human judgements. Moreover, it is five times more memory efficient thanks to its large 200ms units. In addition, the embeddings before and after the Lexical Embedder are phonetically and semantically interpretable.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.