OpenMOSS-Team

/

MOSS-TTS-Realtime

Like 109

Follow

OpenMOSS 1.23k

Hi, I'm Niels from the Hugging Face team.

This model is part of the MOSS-TTS family, which builds upon the CAT architecture described in the recent paper: MOSS-Audio-Tokenizer: Scaling Audio Tokenizers for Future Audio Foundation Models.

I've opened this PR to:

• Update the Arxiv badge from "Coming soon" to point to the actual technical report.
• Add library_name: transformers to the metadata (since it's a custom model compatible with the library).
• Set the pipeline_tag to text-to-speech for improved discoverability.
• Add a reference to the research paper.

These changes help users find the research context and improve the model's visibility on the Hub.