DanielMarius/model_1.safetensors

model 1

This repository contains one safetensors artifact from the Daniel Media Studio archive. It is separated so the file, provenance, license status, and usage notes remain reviewable on one page.

Artifact

• Original file: 'model_1.safetensors'
• Size: 69768280 bytes
• SHA-256: '850799d78699134b969056183fc9d490c51f8d8154d0ed00fae3e738b6b30af6'
• Role: voice or vocoder component
• Creator or upstream attribution: Not established from local source evidence

How to use it

This is an archive component, not a guaranteed standalone pipeline. Select the compatible DMS workflow, place the file in the model directory expected by that workflow, and load any required base model, VAE, tokenizer, encoder, ControlNet, or companion files. LoRA and adapter files require a compatible base pipeline. Training-state files are not inference models.

Required companions

• Use the compatible DMS workflow and the companion files named in source/.
• Check the original source documentation before production or commercial use.

Provenance

Source repositories

• See preserved source URLs

Source URLs

• https://img.shields.io/badge/arXiv-2508.16332-brightgreen.svg?style=flat-square
• https://arxiv.org/abs/2508.16332
• https://img.shields.io/badge/%F0%9F%A4%97%20HuggingFace-model-yellow
• https://huggingface.co/RMSnow/Vevo2
• https://img.shields.io/badge/WebPage-Demo-red.svg
• https://versasinger.github.io/
• https://img.shields.io/badge/README-GitHub-blue
• https://github.com/open-mmlab/Amphion/blob/main/models/svc/vevo2/README.md
• https://huggingface.co/RMSnow/Vevo2/tree/main/tokenizer/prosody_fvq512_6.25hz
• https://huggingface.co/RMSnow/Vevo2/tree/main/tokenizer/contentstyle_fvq16384_12.5hz
• https://huggingface.co/RMSnow/Vevo2/tree/main/contentstyle_modeling/posttrained
• https://huggingface.co/RMSnow/Vevo2/tree/main/acoustic_modeling/fm_emilia101k_singnet7k_repa
• https://huggingface.co/RMSnow/Vevo2/tree/main/vocoder
• https://openreview.net/pdf?id=X6ffdf6nh3

License and attribution

License is unknown from available evidence. This archive grants no new license; verify upstream terms before reuse or redistribution.

The source/ directory preserves local README, source-note, metadata, workflow, license, and small preview evidence when available. No new ownership claim is made over upstream models or materials.