---
language:
- zh
pipeline_tag: text-to-speech
tags:
- A-SOUL
- asoul
- jiaran
- VirtualIdol
- Vtuber
---

This is a model library based on **SO-VITS-SVC4.1** for **A-SOUL(嘉然&贝拉&乃琳)** and members of **Shining Stage(思诺&心宜)** (**Zhijiang Entertainment**).

SO-VITS-SVC4.1 is adopted from the project: [SO-VITS-SVC](https://github.com/svc-develop-team/so-vits-svc).

Thanks to the author, 羽毛布団 (bilibili id: 3493141443250876), for the project integration package.

The training set is sourced from Baidu Net Disk of "奶粉_day0" (bilibili id: 5273959), thanks for his organization.

Each role is trained using audio from four solo live broadcasts.

# Models

## Includes 3 single-speaker models:

嘉然Diana(speaker=jiaran): G_72000.pth

心宜 Fiona(speaker=xinyi): G_92400.pth

思诺Gladys(speaker=sinuo): G_74400.pth

## 1 multi-speaker model:

贝拉Bella&乃琳Eileen(speaker0=beila, speaker1=nailin): G_65600.pth

# To the five girls:

嘉然今天吃什么：https://space.bilibili.com/672328094

贝拉kira：https://space.bilibili.com/672353429

乃琳Queen：https://space.bilibili.com/672342685

心宜不是心仪：https://space.bilibili.com/3537115310721181

思诺snow：https://space.bilibili.com/3537115310721781

未来逐步更新：向晚Ava 和 珈乐Carol（和 莱芙Laffey？（和 阿花？））