espnet

/

visinger2-zh-jp-multisinger-svs

like 0

Follow

ESPnet 322

This PR significantly enhances the model card for the espnet/visinger2-zh-jp-multisinger-svs model by:

• Adding the pipeline_tag: text-to-audio to improve discoverability for users looking for singing voice synthesis models.
• Including the license: mit based on the explicit license information found in the associated GitHub repository.
• Providing direct links to the scientific paper (SingingSDS: A Singing-Capable Spoken Dialogue System for Conversational Roleplay Applications), the project's GitHub repository (https://github.com/SingingSDS/SingingSDS), and the live Hugging Face demo (https://huggingface.co/spaces/espnet/SingingSDS).
• Adding a descriptive title and an overview with a visual element (demo image) for better clarity and engagement.

These additions make the model card more informative and accessible, helping users understand the model's purpose, usage, and context more effectively.