Laurent Mazare 08ba5dae03 Initial commit
2025-06-16 21:36:53 +02:00
2025-06-16 21:36:53 +02:00
2025-06-16 21:36:53 +02:00

delayed-streams-modeling

Delayed Streams Modeling (DSM) is a flexible formulation for streaming, multimodal sequence-to-sequence learning.

Description
Kyutai's Speech-To-Text and Text-To-Speech models based on the Delayed Streams Modeling framework.
Readme 1.9 MiB
Languages
Python 79.3%
Jupyter Notebook 11.6%
Rust 9.1%