Commit
·
35879fa
1
Parent(s):
4f9b2a2
Add README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
tags:
|
| 5 |
+
- rvc
|
| 6 |
+
- voice cloning
|
| 7 |
+
- The Amazing World of Gumball
|
| 8 |
+
- Anais Watterson
|
| 9 |
+
- Kyla Rae Kowalewski
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
## Model Details
|
| 13 |
+
|
| 14 |
+
Voice of Kyla Rae Kowalewski as Anais Watterson in Season 3 of the cartoon The Amazing World of Gumball.
|
| 15 |
+
|
| 16 |
+
### Model Description
|
| 17 |
+
|
| 18 |
+
<!-- Provide a longer summary of what this model is. -->
|
| 19 |
+
|
| 20 |
+
- **Developed by:** [ijik-loker](https://huggingface.co/ijik-loker)
|
| 21 |
+
- **Model type:** [Retrieval-based Voice Conversion (RVC)](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI)
|
| 22 |
+
- **Language(s):** English
|
| 23 |
+
|
| 24 |
+
## Uses
|
| 25 |
+
|
| 26 |
+
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
|
| 27 |
+
|
| 28 |
+
Used in the popular Retrieval-based Voice Conversion WebUI via inference or real-time using [Voice Changer](https://github.com/w-okada/voice-changer).
|
| 29 |
+
The index file should be used alongside the model.
|
| 30 |
+
|
| 31 |
+
## Training Details
|
| 32 |
+
|
| 33 |
+
### Training Data
|
| 34 |
+
|
| 35 |
+
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
|
| 36 |
+
|
| 37 |
+
#### Voice clips dataset total duration
|
| 38 |
+
|
| 39 |
+
v1 model: 03min 56s
|
| 40 |
+
|
| 41 |
+
Trained using these episodes from Season 3:
|
| 42 |
+
1. The Nobody
|
| 43 |
+
2. The Name
|
| 44 |
+
3. The Oracle
|
| 45 |
+
4. The Safety
|
| 46 |
+
5. The Procrastinators
|
| 47 |
+
6. The Puppy
|
| 48 |
+
7. The Spoiler
|
| 49 |
+
|
| 50 |
+
### Training Procedure
|
| 51 |
+
|
| 52 |
+
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
| 53 |
+
|
| 54 |
+
1. Remove noise using [Ultimate Vocal Remover 5](https://github.com/Anjok07/ultimatevocalremovergui) UVR-DeNoise.
|
| 55 |
+
2. Extract vocals using RVC Web UI [HP5-主旋律人声vocals+其他instrumentals.pth](https://huggingface.co/lj1995/VoiceConversionWebUI/blob/main/uvr5_weights/HP5-%E4%B8%BB%E6%97%8B%E5%BE%8B%E4%BA%BA%E5%A3%B0vocals%2B%E5%85%B6%E4%BB%96instrumentals.pth).
|
| 56 |
+
3. Remove echo and reverb using Ultimate Vocal Remover 5 UVR-DeEcho-DeReverb.
|
| 57 |
+
4. Manually diarise voices in [Audacity](https://www.audacityteam.org/) using labels.
|
| 58 |
+
5. Export multiple to .wav by labels.
|
| 59 |
+
6. Train using RVC
|
| 60 |
+
* Target Sample Rate: 48k
|
| 61 |
+
* Version: v2
|
| 62 |
+
* Total training epochs: 200
|
| 63 |
+
* Base model G: f0G48k.pth
|
| 64 |
+
* Base model D: f0D48k.pth
|