DanielMarius/dms-vevo2-whisper-medium

Vevo2 Whisper medium

This Hugging Face model repository is a named bundle from the Drakula Media Studio project.

• Model name: Vevo2 Whisper medium
• Artifact family: audio/voice component
• License: UNKNOWN
• Payload files: 1
• Payload size: 1528008539 bytes

The repository name follows the model or model family. Generic files such as pytorch_model.bin, model.pt, model.safetensors, or .bin are payload filenames inside this repository, not separate repositories.

Provenance

The files were collected from the local DMS/Drakula model stores. Source paths and candidate content identities are recorded in DMS-MODEL-NOTES.md.

Usage

Use the files according to the upstream model family and the directory names preserved in this repository. Check the upstream project documentation before redistribution or production use.

Vevo2 Whisper statistics companion

Adds the local Whisper statistics tensor required by the Vevo2 singing-voice configuration, and documents its Amphion upstream. The file remains a configuration companion, not a separate model.

Files and local identity

• config/whisper_stats.pt — 9215 bytes; SHA-256 6117052b3e23e9075a79cc208ecc24328ae2f71a7dd4c9793db7c90a88b4a519

Upstream copies and provenance

• Related upstream: amphion/Vevo.

License and use

License: UNKNOWN. No reliable license was found for this local artifact. This repository grants no additional rights; verify upstream terms before reuse or redistribution.

Provenance audit

Generated from local DMS model assets. Audit run: dms-model-assets-publish-20261022T0750Z-v7. SHA-256 values identify the local source bytes.