Datasets:
Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
metrics:
|
| 6 |
+
- wer
|
| 7 |
+
pipeline_tag: audio-to-audio
|
| 8 |
+
tags:
|
| 9 |
+
- rvc
|
| 10 |
+
- audio
|
| 11 |
+
- speech
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
|
| 15 |
+
# Obama Voice Sample Dataset for RVC Training
|
| 16 |
+
|
| 17 |
+
A curated dataset of Barack Obama's voice samples, specifically prepared for training Demo RVC (Retrieval-based Voice Conversion) model on ModelsLab.
|
| 18 |
+
|
| 19 |
+
## Dataset Specifications
|
| 20 |
+
- Total Duration: 25+ minutes
|
| 21 |
+
- Audio Format: WAV
|
| 22 |
+
- Sampling Rate: 24 kHz
|
| 23 |
+
- Content Type: Clean speech samples from speeches and addresses
|
| 24 |
+
|
| 25 |
+
## Usage
|
| 26 |
+
This dataset is designed for training RVC (Retrieval-based Voice Conversion) models. The minimum recommended training duration for RVC is 25 minutes of clean audio, which this dataset provides.
|
| 27 |
+
|
| 28 |
+
## Important Notes
|
| 29 |
+
- Use this dataset responsibly and in accordance with ethical AI guidelines
|
| 30 |
+
- The dataset is intended for research and educational purposes
|
| 31 |
+
- Please check local regulations regarding voice cloning and impersonation
|
| 32 |
+
|
| 33 |
+
## Source
|
| 34 |
+
Audio samples are collected from public speeches and addresses given by Barack Obama.
|
| 35 |
+
|
| 36 |
+
## Requirements for RVC Training
|
| 37 |
+
- Minimum audio length: 25 minutes
|
| 38 |
+
- Format: WAV files
|
| 39 |
+
- Sample rate: 24 kHz (minimum)
|
| 40 |
+
- Clean audio without background noise
|