File size: 6,203 Bytes
bc65357
 
 
 
 
 
 
 
 
 
 
442cf63
bc65357
 
 
 
 
 
d25ce06
bc65357
 
78aaf30
 
442cf63
bc65357
442cf63
 
 
944c4e8
442cf63
 
 
 
 
 
90b171f
442cf63
4a28762
442cf63
944c4e8
442cf63
 
4a28762
 
944c4e8
4a28762
 
ebc191c
 
17758ed
 
ebc191c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
17758ed
 
44b1458
 
2b07646
 
 
 
ebc191c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bc65357
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
---
language:
- zh
tags:
- AIvtuber
- VirtuaReal
---
# SUImodels

### 岁己所有的模型都在这里

### 包括sovits3.0、4.0及onnx,还有以后会出的vits模型

## Model Description

<!-- Provide a longer summary of what this model is. -->

- **Developed by:** [謬紗特](https://space.bilibili.com/59442895)
- **Model type:** [so-vits-svc 3.0 48kHz](https://github.com/innnky/so-vits-svc/tree/main)、[so-vits-svc 4.0](https://github.com/innnky/so-vits-svc/tree/4.0)
- **Demo:** [SPACE: AI岁己(歌声变声器)](https://huggingface.co/spaces/Miuzarte/SUI-svc-3.0)

### pth文件名的训练步数是程序按学习率等超参数得出的步数,onnx文件名的步数为实际训练步数

  |Sovits3_v1|Base/G_1000000.pth|Singing/G_1M111000.pth|Singing/G_100000.pth|
  |-:|:-:|:-:|:-:|
  |onnx|Base/suijiSUI_v1_1M_SoVits.onnx|Singing/suijiSUI_v1_1M111000_SoVits.onnx|Singing/suijiSUI_v1_100000_SoVits.onnx|
  |训练集|12月录播(除电台)、出道至今22条歌投、10条歌切、圣诞音声(27.5小时)|Base/G_1000000.pth作为底模_2022年所有唱歌投稿、唱歌切片、圣诞音声(3.9小时)|2022年所有唱歌投稿、唱歌切片、圣诞音声(3.9小时)|

### 因为v2练着练着突然sovits4.0就出来了所以200k直接收了,弃用

  |Sovits3_v2|Base/G_100000.pth|Singing/G_160000.pth|
  |-:|:-:|:-:|
  |onnx|Base/suijiSUI_v2_100000_SoVits.onnx|Singing/suijiSUI_v2_100k100000_SoVits.onnx|
  |训练集|22年12月、23年1月的录播(06:47:46)|Base/G_100000.pth作为底模_22年12月、23年1月、23年2月1-17日的录播(除电台,共计268:07:43)、岁己的投稿、A1in_sy11月及以前的歌切|

### Singing/G_160000.pth、Singing/G_210000.pth、Singing/G_260000.pth我个人倾向于折中使用Singing/G_210000.pth

  |Sovits4_v3|Base/G_100000.pth|Singing/G_160000.pth|
  |-:|:-:|:-:|
  |onnx|Base/suijiSUI_v3_100000_SoVits.onnx|Singing/suijiSUI_v3_100k100000_SoVits.onnx|
  |训练集|22年12月、23年1月的录播(06:47:46)|Base/G_100000.pth作为底模_22年12月、23年1月、23年2月1-17日的录播(除电台,共计268:07:43)、岁己的投稿、A1in_sy11月及以前的歌切|

  |Sovits4_v3|Singing/G_210000.pth|Singing/G_260000.pth|备注|
  |-:|:-:|:-:|:-:|
  |onnx|Singing/suijiSUI_v3_100k150000_SoVits.onnx|Singing/suijiSUI_v3_100k200000_SoVits.onnx|160k开始loss就没再往下了,这两个估计有一丁点过拟合,然后我个人|
  |训练集|{同Singing/G_160000.pth}|{同Singing/G_160000.pth}|也听不出这三个有什么区别,有强迫症的可以自己再仔细对比一下|

## MoeSS\\Mods配置文件

#### sovits3.0需要MoeSS\\hubert\\[hubert.onnx](https://huggingface.co/NaruseMioShirakana/MoeSS-SUBModel/blob/main/hubert.7z)

岁己SUI_v1_1M.json (suijiSUI_v1_1M\\)

```json
{
  "Folder" : "suijiSUI_v1_1M",
  "Name" : "岁己SUI_v1_1M",
  "Type" : "SoVits",
  "Rate" : 48000,
  "Hop" : 320,
  "Hubert": "hubert",
  "SoVits3": true,
  "Characters" : ["岁己SUI"]
}
```

岁己SUI_v1_1M111k.json (suijiSUI_v1_1M111000\\)

```json
{
  "Folder" : "suijiSUI_v1_1M111000",
  "Name" : "岁己SUI_v1_1M111k",
  "Type" : "SoVits",
  "Rate" : 48000,
  "Hop" : 320,
  "Hubert": "hubert",
  "SoVits3": true,
  "Characters" : ["岁己SUI"]
}
```

岁己SUI_v1_100k.json (suijiSUI_v1_1M\\)

```json
{
  "Folder" : "suijiSUI_v1_1M",
  "Name" : "岁己SUI_v1_1M",
  "Type" : "SoVits",
  "Rate" : 48000,
  "Hop" : 320,
  "Hubert": "hubert",
  "SoVits3": true,
  "Characters" : ["岁己SUI"]
}
```

岁己SUI_v2_100k.json (suijiSUI_v2_100000\\)

```json
{
  "Folder" : "suijiSUI_v2_100000",
  "Name" : "岁己SUI_v2_100k",
  "Type" : "SoVits",
  "Rate" : 48000,
  "Hop" : 320,
  "Hubert": "hubert",
  "SoVits3": true,
  "Characters" : ["岁己SUI"]
}
```

岁己SUI_v2_100k100k.json (suijiSUI_v2_100k100000\\)

```json
{
  "Folder" : "suijiSUI_v2_100k100000",
  "Name" : "岁己SUI_v2_100k100k",
  "Type" : "SoVits",
  "Rate" : 48000,
  "Hop" : 320,
  "Hubert": "hubert",
  "SoVits3": true,
  "Characters" : ["岁己SUI"]
}
```

#### sovits4.0需要MoeSS\\hubert\\[hubert4.0.onnx](https://huggingface.co/NaruseMioShirakana/MoeSS-SUBModel/blob/main/hubert4.0.7z)

sovits4.0被支持于MoeSS v4.2.0,建议使用最新版,作者已将仓库存档[[MoeSS/releases]](https://github.com/NaruseMioShirakana/MoeSS/releases)

#### 更建议使用[sovits4.0的inference_main.py](https://github.com/innnky/so-vits-svc/tree/4.0)进行推理

之后更新聚类模型

岁己SUI_v3_100k.json (suijiSUI_v3_100000\\)

```json
{
  "Folder" : "suijiSUI_v3_100000",
  "Name" : "岁己SUI_v3_100k",
  "Type" : "SoVits",
  "Rate" : 44100,
  "Hop" : 512,
  "Hubert": "hubert4.0",
  "SoVits4": true,
  "Characters" : ["岁己SUI"]
}
```

岁己SUI_v3_100k100k.json (suijiSUI_v3_100k100000\\)

```json
{
  "Folder" : "suijiSUI_v3_100k100000",
  "Name" : "岁己SUI_v3_100k100k",
  "Type" : "SoVits",
  "Rate" : 44100,
  "Hop" : 512,
  "Hubert": "hubert4.0",
  "SoVits4": true,
  "Characters" : ["岁己SUI"]
}
```

岁己SUI_v3_100k150k.json (suijiSUI_v3_100k150000\\)

```json
{
  "Folder" : "suijiSUI_v3_100k150000",
  "Name" : "岁己SUI_v3_100k150k",
  "Type" : "SoVits",
  "Rate" : 44100,
  "Hop" : 512,
  "Hubert": "hubert4.0",
  "SoVits4": true,
  "Characters" : ["岁己SUI"]
}
```

岁己SUI_v3_100k200k.json (suijiSUI_v3_100k200000\\)

```json
{
  "Folder" : "suijiSUI_v3_100k200000",
  "Name" : "岁己SUI_v3_100k200k",
  "Type" : "SoVits",
  "Rate" : 44100,
  "Hop" : 512,
  "Hubert": "hubert4.0",
  "SoVits4": true,
  "Characters" : ["岁己SUI"]
}
```

# Uses

<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->

## Direct Use

<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->

[More Information Needed]

## Downstream Use [optional]

<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->

[More Information Needed]

## Out-of-Scope Use

<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->

[More Information Needed]