Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
datasets:
|
| 3 |
+
- simon3000/genshin-voice
|
| 4 |
+
- CSTR-Edinburgh/vctk
|
| 5 |
+
language:
|
| 6 |
+
- en
|
| 7 |
+
---
|
| 8 |
+
# So-Vits-Svc Base Model V1
|
| 9 |
+
The base model to generate new voices with so-vits-svc voice lab.
|
| 10 |
+
|
| 11 |
+
The dataset was comprised of 278 english speaking people.
|
| 12 |
+
4 datasets where used:
|
| 13 |
+
- Genshin Voice: Only speakers with more than 30min of audio
|
| 14 |
+
- VCTK
|
| 15 |
+
- Vocalset
|
| 16 |
+
- Private scraped dataset
|
| 17 |
+
|
| 18 |
+
The model was trained for around 4 days and 16 hours on a single rtx 3090 (61 epochs / 430k steps)
|
| 19 |
+
|