Datasets:
Update README_EN.md
Browse files- README_EN.md +126 -0
README_EN.md
CHANGED
@@ -0,0 +1,126 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
|
2 |
+
|
3 |
+
This repository is used to preview the effects of various speech models trained by so-vits-svc-4.0.
|
4 |
+
**Click on the character name** to automatically jump to the corresponding training parameters.</br>
|
5 |
+
I recommend using **Google Chrome** as other browsers may not load the previewed audio correctly.</br>
|
6 |
+
The conversion of normal speaking voices is relatively accurate, but songs with a wide range of sounds
|
7 |
+
and background music and other noises that are difficult to remove may result in a unstable effect.</br>
|
8 |
+
If you have recommended songs that you would like to try converting and listening to or any other suggestions,
|
9 |
+
[**click here**](https://huggingface.co/datasets/jiaheillu/audio_preview/discussions/new) to give me advice.</br>
|
10 |
+
Below are preview audios. **Scroll up, down, left, and right** to see them all.
|
11 |
+
|
12 |
+
<style>
|
13 |
+
.scrolling-container {
|
14 |
+
width: 100%;
|
15 |
+
max-width: 800px;
|
16 |
+
height: 300px;
|
17 |
+
overflow: auto;
|
18 |
+
margin: 0;
|
19 |
+
}
|
20 |
+
@media screen and (max-width: 768px) {
|
21 |
+
.scrolling-container {
|
22 |
+
width: 100%;
|
23 |
+
height: auto;
|
24 |
+
overflow: auto;
|
25 |
+
}
|
26 |
+
}
|
27 |
+
</style>
|
28 |
+
|
29 |
+
<div class="scrolling-container">
|
30 |
+
<table border="1" style="white-space: nowrap; text-align: center;">
|
31 |
+
<thead>
|
32 |
+
<tr>
|
33 |
+
<th>Character Name</th>
|
34 |
+
<th>Original Voice A</th>
|
35 |
+
<th>Converted Voice B</th>
|
36 |
+
<th>A Voice Replaced by B</th>
|
37 |
+
<th>Song Cover (Click to Download)</th>
|
38 |
+
</tr>
|
39 |
+
</thead>
|
40 |
+
<tbody>
|
41 |
+
<tr>
|
42 |
+
<td><a href="https://huggingface.co/datasets/jiaheillu/audio_preview/blob/main/散兵效果预览/训练参数速览.md">Wanderer</a></td>
|
43 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/散兵效果预览/部分训练集/真遗憾,小吉祥草王让他消除了那么多的切片,剥夺了我将他一片一片千刀万剐的快乐%E3%80%82.mp3" controls="controls"></audio></td>
|
44 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/散兵效果预览/原声/shenli3.wav" controls="controls"></audio></td>
|
45 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/散兵效果预览/转换结果/shenli3mp3_auto_liulangzhe.wav" controls="controls"></audio></td>
|
46 |
+
<td><a href="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/散兵效果预览/转换结果/夢で逢えたら2liulangzhe_f.wav">夢で会えたら</a></td>
|
47 |
+
</tr>
|
48 |
+
<tr>
|
49 |
+
<td><a href="https://huggingface.co/datasets/jiaheillu/audio_preview/blob/main/胡桃_preview/README.md">HuTao</a></td>
|
50 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/%E8%83%A1%E6%A1%83_preview/hutao.wav" controls="controls"></audio></td>
|
51 |
+
<td>.........</td>
|
52 |
+
<td>.........</td>
|
53 |
+
<td>
|
54 |
+
<a href="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/胡桃_preview/moonlight_shadow2胡桃.WAV">moonlight shadow</a>,
|
55 |
+
<a href="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/胡桃_preview/云烟成雨2胡桃.WAV">云烟成雨</a>,
|
56 |
+
<a href="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/胡桃_preview/原点2胡桃.WAV">原点</a>,
|
57 |
+
<a href="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/胡桃_preview/夢だ会えたら2胡桃.WAV">夢で逢えたら</a>,
|
58 |
+
<a href="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/胡桃_preview/贝加尔湖畔2胡桃.WAV">贝加尔湖畔</a>
|
59 |
+
</td>
|
60 |
+
</tr>
|
61 |
+
<tr>
|
62 |
+
<td><a href="https://huggingface.co/datasets/jiaheillu/audio_preview/blob/main/绫华_preview/README.md">Kamisato Ayaka</a></td>
|
63 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/绫华_preview/linghua428.wav" controls="controls"></audio></td>
|
64 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/绫华_preview/yelan.wav" controls="controls"></audio></td>
|
65 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/绫华_preview/yelan.wav_auto_linghua_0.5.wav" controls="controls"></audio></td>
|
66 |
+
<td>
|
67 |
+
<a href="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/绫华_preview/アムリタ2绫华.WAV">アムリタ</a>,
|
68 |
+
<a href="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/绫华_preview/大鱼2绫华.WAV">大鱼</a>,
|
69 |
+
<a href="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/绫华_preview/遊園施設2绫华.WAV">遊園施設</a>,
|
70 |
+
<a href="https://huggingface.co/datasets/jiaheillu/audio_preview/resolve/main/绫华_preview/the_day_you_want_away2绫华.WAV">the day you want away</a>
|
71 |
+
</td>
|
72 |
+
</tr>
|
73 |
+
<tr>
|
74 |
+
<td><a href="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/blob/main/宵宫_preview/README.md">yoimiya</a></td>
|
75 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/宵宫_preview/xiaogong.wav" controls="controls"></audio></td>
|
76 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/宵宫_preview/hutao2.wav" controls="controls"></audio></td>
|
77 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/宵宫_preview/hutao2wav_0key_xiaogong_0.5-2.wav" controls="controls"></audio></td>
|
78 |
+
<td>
|
79 |
+
<a href="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/宵宫_preview/昨夜书2宵宫.WAV">昨夜书</a>,
|
80 |
+
<a href="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/宵宫_preview/lemon2宵宫.WAV">lemon</a>,
|
81 |
+
<a href="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/宵宫_preview/my_heart_will_go_no2宵宫.WAV">my heart will go on</a>,
|
82 |
+
</td>
|
83 |
+
</tr>
|
84 |
+
<tr>
|
85 |
+
<td><a href="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/blob/main/刻晴_preview/README.md">Keqing</a></td>
|
86 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/刻晴_preview/原_keqing2.wav" controls="controls"></audio></td>
|
87 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/刻晴_preview/待_xiaogong3.wav" controls="controls"></audio></td>
|
88 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/刻晴_preview/已_xiaogong2keqing.wav" controls="controls"></audio></td>
|
89 |
+
<td>
|
90 |
+
<a href="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/刻晴_preview/嚣张2刻晴.WAV">嚣张</a>,
|
91 |
+
<a href="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/刻晴_preview/ファティマ2刻晴.WAV">ファティマ</a>,
|
92 |
+
<a href="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/刻晴_preview/hero2刻晴.WAV">hero</a>,
|
93 |
+
</td>
|
94 |
+
</tr>
|
95 |
+
<tr>
|
96 |
+
<td><a href="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/blob/main/imallryt_preview/README.md">imallryt</a></td>
|
97 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/imallryt_preview/%E5%8E%9F_IVOL_1%20Care_DRY_120_Am_Main_Vocal.wav" controls="controls"></audio></td>
|
98 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/imallryt_preview/%E5%BE%85_Lead_A%20minor_DRY.wav" controls="controls"></audio></td>
|
99 |
+
<td><audio src="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/imallryt_preview/%E5%B7%B2_Lead_A%20minor_DRYwav_0key_imallryt_0.5.wav" controls="controls"></audio></td>
|
100 |
+
<td>
|
101 |
+
<a href="https://huggingface.co/datasets/jiaheillu/sovits_audio_preview/resolve/main/imallryt_preview/海阔天空2imallryt.WAV">海阔天空</a>,
|
102 |
+
</td>
|
103 |
+
</tr>
|
104 |
+
</tbody>
|
105 |
+
</table>
|
106 |
+
</div>
|
107 |
+
|
108 |
+
Key Parameters:
|
109 |
+
|
110 |
+
audio duration: total duration of the training dataset </br>
|
111 |
+
|
112 |
+
epoch: number of rounds of training</br>
|
113 |
+
|
114 |
+
Others: </br>
|
115 |
+
|
116 |
+
batch_size = number of audio segments trained in one step </br>
|
117 |
+
|
118 |
+
segments = the number of segments that the audio is split into ,step = segments * epoch / batch_size, which is where the numbers in the model file name come from
|
119 |
+
|
120 |
+
Using the example of "Wanderer" (a character name): Loss Function Graph: pay attention to step and loss5,
|
121 |
+
for example:<br> As a difficult test, all the original audios are high-pitched female voices, and this graph
|
122 |
+
shows the result of training on a 10-minute pure voice audio. The model achieved good performance at around
|
123 |
+
2800 epochs (10,000 steps), and the actual model used was trained for 5571 epochs (19,500 steps), with
|
124 |
+
slight differences between the trained voice and the original voice. Please refer to the preview audio above.
|
125 |
+
In general, 10 minutes is not enough for a sufficient training dataset.
|
126 |
+
[Click here to view related files](https://huggingface.co/datasets/jiaheillu/audio_preview/tree/main)<br>
|