Update README.md
Browse files
README.md
CHANGED
|
@@ -1,187 +1,187 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
language:
|
| 4 |
-
- en
|
| 5 |
-
tags:
|
| 6 |
-
- ai
|
| 7 |
-
- rvc
|
| 8 |
-
- vc
|
| 9 |
-
- voice-cloning
|
| 10 |
-
- applio
|
| 11 |
-
- titan
|
| 12 |
-
- pretrained
|
| 13 |
-
datasets:
|
| 14 |
-
- blaise-tk/TITAN-Medium
|
| 15 |
-
pipeline_tag: audio-to-audio
|
| 16 |
-
---
|
| 17 |
-
|
| 18 |
-
# TITAN: A Versatile, Robust, and High-Quality Pretrained Model for Retrieval-based Voice Conversion (RVC) Training
|
| 19 |
-
|
| 20 |
-
## Overview
|
| 21 |
-
|
| 22 |
-
TITAN is a state-of-the-art pretrained model designed for Retrieval-based Voice Conversion (https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/) training. It offers a robust solution for transforming voice characteristics from one speaker to another, providing high-quality results with minimal training effort.
|
| 23 |
-
|
| 24 |
-
## Model Details
|
| 25 |
-
|
| 26 |
-
### Titan-Medium
|
| 27 |
-
|
| 28 |
-
- Training Environment: Utilized a RTX 3060 TI on Applio v3.1.1 (https://github.com/IAHispano/Applio), employing a batch size of 8 over a span of 3 weeks.
|
| 29 |
-
- Iterations (48k): 1018660 Steps and 530 Epochs
|
| 30 |
-
- Iterations (40k): 1010588 Steps and 467 Epochs
|
| 31 |
-
- Iterations (32k): 1001469 Steps and 463 Epochs
|
| 32 |
-
- Sampling rate: 48k
|
| 33 |
-
- Fine-tuning Process: RVC v2 pretrained with pitch guidance, leveraging an 11.15-hour dataset sourced from Expresso (https://arxiv.org/abs/2308.05725) also available on [datasets/blaise-tk/TITAN-Medium](https://huggingface.co/datasets/blaise-tk/TITAN-Medium).
|
| 34 |
-
|
| 35 |
-
#### Samples
|
| 36 |
-
*Tests performed with a premature ckpt at ~700k steps doing all tests under the same conditions.*
|
| 37 |
-
|
| 38 |
-
<table style="width:100%; text-align:center;">
|
| 39 |
-
<tr>
|
| 40 |
-
<th>Titan-Medium</th>
|
| 41 |
-
<th>Ov2</th>
|
| 42 |
-
<th>Ov2.1</th>
|
| 43 |
-
</tr>
|
| 44 |
-
<tr>
|
| 45 |
-
<td>
|
| 46 |
-
<audio controls>
|
| 47 |
-
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 1 - Test 1 - Titan.wav?download=true" type="audio/wav">
|
| 48 |
-
Your browser does not support the audio element.
|
| 49 |
-
</audio>
|
| 50 |
-
</td>
|
| 51 |
-
<td>
|
| 52 |
-
<audio controls>
|
| 53 |
-
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 1 - Test 1 - Ov2.wav?download=true" type="audio/wav">
|
| 54 |
-
Your browser does not support the audio element.
|
| 55 |
-
</audio>
|
| 56 |
-
</td>
|
| 57 |
-
</tr>
|
| 58 |
-
|
| 59 |
-
</tr>
|
| 60 |
-
<tr>
|
| 61 |
-
<td>
|
| 62 |
-
<audio controls>
|
| 63 |
-
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 1 - Test 2 - Titan.wav?download=true" type="audio/wav">
|
| 64 |
-
Your browser does not support the audio element.
|
| 65 |
-
</audio>
|
| 66 |
-
</td>
|
| 67 |
-
<td>
|
| 68 |
-
<audio controls>
|
| 69 |
-
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 1 - Test 2 - Ov2.wav?download=true" type="audio/wav">
|
| 70 |
-
Your browser does not support the audio element.
|
| 71 |
-
</audio>
|
| 72 |
-
</td>
|
| 73 |
-
</tr>
|
| 74 |
-
|
| 75 |
-
<tr>
|
| 76 |
-
<td>
|
| 77 |
-
<audio controls>
|
| 78 |
-
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 2 - Test 1 - Titan.wav?download=true" type="audio/wav">
|
| 79 |
-
Your browser does not support the audio element.
|
| 80 |
-
</audio>
|
| 81 |
-
</td>
|
| 82 |
-
<td>
|
| 83 |
-
<audio controls>
|
| 84 |
-
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 2 - Test 1 - Ov2.wav?download=true" type="audio/wav">
|
| 85 |
-
Your browser does not support the audio element.
|
| 86 |
-
</audio>
|
| 87 |
-
</td>
|
| 88 |
-
|
| 89 |
-
</tr>
|
| 90 |
-
<tr>
|
| 91 |
-
<td>
|
| 92 |
-
<audio controls>
|
| 93 |
-
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 2 - Test 2 - Titan.wav?download=true" type="audio/wav">
|
| 94 |
-
Your browser does not support the audio element.
|
| 95 |
-
</audio>
|
| 96 |
-
</td>
|
| 97 |
-
<td>
|
| 98 |
-
<audio controls>
|
| 99 |
-
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 2 - Test 2 - Ov2.wav?download=true" type="audio/wav">
|
| 100 |
-
Your browser does not support the audio element.
|
| 101 |
-
</audio>
|
| 102 |
-
</td>
|
| 103 |
-
</tr>
|
| 104 |
-
|
| 105 |
-
</tr>
|
| 106 |
-
<tr>
|
| 107 |
-
<td>
|
| 108 |
-
<audio controls>
|
| 109 |
-
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 3 - Test 1 - Titan.wav?download=true" type="audio/wav">
|
| 110 |
-
Your browser does not support the audio element.
|
| 111 |
-
</audio>
|
| 112 |
-
</td>
|
| 113 |
-
<td>
|
| 114 |
-
<audio controls>
|
| 115 |
-
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 3 - Test 1 - Ov2.wav?download=true" type="audio/wav">
|
| 116 |
-
Your browser does not support the audio element.
|
| 117 |
-
</audio>
|
| 118 |
-
</td>
|
| 119 |
-
<td>
|
| 120 |
-
<audio controls>
|
| 121 |
-
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 3 - Test 1 - Ov2.1.wav?download=true" type="audio/wav">
|
| 122 |
-
Your browser does not support the audio element.
|
| 123 |
-
</audio>
|
| 124 |
-
</td>
|
| 125 |
-
</tr>
|
| 126 |
-
|
| 127 |
-
</tr>
|
| 128 |
-
<tr>
|
| 129 |
-
<td>
|
| 130 |
-
<audio controls>
|
| 131 |
-
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 3 - Test 2 - Titan.wav?download=true" type="audio/wav">
|
| 132 |
-
Your browser does not support the audio element.
|
| 133 |
-
</audio>
|
| 134 |
-
</td>
|
| 135 |
-
<td>
|
| 136 |
-
<audio controls>
|
| 137 |
-
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 3 - Test 2 - Ov2.wav?download=true" type="audio/wav">
|
| 138 |
-
Your browser does not support the audio element.
|
| 139 |
-
</audio>
|
| 140 |
-
</td>
|
| 141 |
-
<td>
|
| 142 |
-
<audio controls>
|
| 143 |
-
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 3 - Test 2 - Ov2.1.wav?download=true" type="audio/wav">
|
| 144 |
-
Your browser does not support the audio element.
|
| 145 |
-
</audio>
|
| 146 |
-
</td>
|
| 147 |
-
</tr>
|
| 148 |
-
|
| 149 |
-
</table>
|
| 150 |
-
|
| 151 |
-
### Titan-Large
|
| 152 |
-
|
| 153 |
-
- Details forthcoming...
|
| 154 |
-
|
| 155 |
-
## Collaborators
|
| 156 |
-
|
| 157 |
-
We appreciate the contributions of our collaborators who have helped in the development and refinement of TITAN.
|
| 158 |
-
|
| 159 |
-
- Mustar
|
| 160 |
-
- SimplCup
|
| 161 |
-
- UnitedShoes
|
| 162 |
-
|
| 163 |
-
## Beta Testers
|
| 164 |
-
|
| 165 |
-
We extend our gratitude to the beta testers who provided valuable feedback during the testing phase of TITAN.
|
| 166 |
-
|
| 167 |
-
- SimplCup
|
| 168 |
-
- Leo_Frixi
|
| 169 |
-
- Light
|
| 170 |
-
- SCRFilms
|
| 171 |
-
- Ryanz
|
| 172 |
-
- Litsa_the_dancer
|
| 173 |
-
|
| 174 |
-
## Citation
|
| 175 |
-
|
| 176 |
-
Should you find TITAN beneficial for your research endeavors or projects, we kindly request citing our repository:
|
| 177 |
-
|
| 178 |
-
```
|
| 179 |
-
@article{titan,
|
| 180 |
-
title={TITAN: A Versatile, Robust, and High-Quality Pretrained Model for Retrieval-based Voice Conversion (RVC) Training},
|
| 181 |
-
author={Blaise},
|
| 182 |
-
journal={Hugging Face},
|
| 183 |
-
year={2024},
|
| 184 |
-
publisher={Blaise},
|
| 185 |
-
url={https://huggingface.co/blaise-tk/TITAN/}
|
| 186 |
-
}
|
| 187 |
-
```
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
tags:
|
| 6 |
+
- ai
|
| 7 |
+
- rvc
|
| 8 |
+
- vc
|
| 9 |
+
- voice-cloning
|
| 10 |
+
- applio
|
| 11 |
+
- titan
|
| 12 |
+
- pretrained
|
| 13 |
+
datasets:
|
| 14 |
+
- blaise-tk/TITAN-Medium
|
| 15 |
+
pipeline_tag: audio-to-audio
|
| 16 |
+
---
|
| 17 |
+
|
| 18 |
+
# TITAN: A Versatile, Robust, and High-Quality Pretrained Model for Retrieval-based Voice Conversion (RVC) Training
|
| 19 |
+
|
| 20 |
+
## Overview
|
| 21 |
+
|
| 22 |
+
TITAN is a state-of-the-art pretrained model designed for Retrieval-based Voice Conversion (https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/) training. It offers a robust solution for transforming voice characteristics from one speaker to another, providing high-quality results with minimal training effort.
|
| 23 |
+
|
| 24 |
+
## Model Details
|
| 25 |
+
|
| 26 |
+
### Titan-Medium
|
| 27 |
+
|
| 28 |
+
- Training Environment: Utilized a RTX 3060 TI on Applio v3.1.1 (https://github.com/IAHispano/Applio), employing a batch size of 8 over a span of 3 weeks.
|
| 29 |
+
- Iterations (48k): 1018660 Steps and 530 Epochs
|
| 30 |
+
- Iterations (40k): 1010588 Steps and 467 Epochs
|
| 31 |
+
- Iterations (32k): 1001469 Steps and 463 Epochs
|
| 32 |
+
- Sampling rate: 48k, 40k, 32k
|
| 33 |
+
- Fine-tuning Process: RVC v2 pretrained with pitch guidance, leveraging an 11.15-hour dataset sourced from Expresso (https://arxiv.org/abs/2308.05725) also available on [datasets/blaise-tk/TITAN-Medium](https://huggingface.co/datasets/blaise-tk/TITAN-Medium).
|
| 34 |
+
|
| 35 |
+
#### Samples
|
| 36 |
+
*Tests performed with a premature ckpt at ~700k steps doing all tests under the same conditions.*
|
| 37 |
+
|
| 38 |
+
<table style="width:100%; text-align:center;">
|
| 39 |
+
<tr>
|
| 40 |
+
<th>Titan-Medium</th>
|
| 41 |
+
<th>Ov2</th>
|
| 42 |
+
<th>Ov2.1</th>
|
| 43 |
+
</tr>
|
| 44 |
+
<tr>
|
| 45 |
+
<td>
|
| 46 |
+
<audio controls>
|
| 47 |
+
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 1 - Test 1 - Titan.wav?download=true" type="audio/wav">
|
| 48 |
+
Your browser does not support the audio element.
|
| 49 |
+
</audio>
|
| 50 |
+
</td>
|
| 51 |
+
<td>
|
| 52 |
+
<audio controls>
|
| 53 |
+
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 1 - Test 1 - Ov2.wav?download=true" type="audio/wav">
|
| 54 |
+
Your browser does not support the audio element.
|
| 55 |
+
</audio>
|
| 56 |
+
</td>
|
| 57 |
+
</tr>
|
| 58 |
+
|
| 59 |
+
</tr>
|
| 60 |
+
<tr>
|
| 61 |
+
<td>
|
| 62 |
+
<audio controls>
|
| 63 |
+
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 1 - Test 2 - Titan.wav?download=true" type="audio/wav">
|
| 64 |
+
Your browser does not support the audio element.
|
| 65 |
+
</audio>
|
| 66 |
+
</td>
|
| 67 |
+
<td>
|
| 68 |
+
<audio controls>
|
| 69 |
+
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 1 - Test 2 - Ov2.wav?download=true" type="audio/wav">
|
| 70 |
+
Your browser does not support the audio element.
|
| 71 |
+
</audio>
|
| 72 |
+
</td>
|
| 73 |
+
</tr>
|
| 74 |
+
|
| 75 |
+
<tr>
|
| 76 |
+
<td>
|
| 77 |
+
<audio controls>
|
| 78 |
+
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 2 - Test 1 - Titan.wav?download=true" type="audio/wav">
|
| 79 |
+
Your browser does not support the audio element.
|
| 80 |
+
</audio>
|
| 81 |
+
</td>
|
| 82 |
+
<td>
|
| 83 |
+
<audio controls>
|
| 84 |
+
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 2 - Test 1 - Ov2.wav?download=true" type="audio/wav">
|
| 85 |
+
Your browser does not support the audio element.
|
| 86 |
+
</audio>
|
| 87 |
+
</td>
|
| 88 |
+
|
| 89 |
+
</tr>
|
| 90 |
+
<tr>
|
| 91 |
+
<td>
|
| 92 |
+
<audio controls>
|
| 93 |
+
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 2 - Test 2 - Titan.wav?download=true" type="audio/wav">
|
| 94 |
+
Your browser does not support the audio element.
|
| 95 |
+
</audio>
|
| 96 |
+
</td>
|
| 97 |
+
<td>
|
| 98 |
+
<audio controls>
|
| 99 |
+
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 2 - Test 2 - Ov2.wav?download=true" type="audio/wav">
|
| 100 |
+
Your browser does not support the audio element.
|
| 101 |
+
</audio>
|
| 102 |
+
</td>
|
| 103 |
+
</tr>
|
| 104 |
+
|
| 105 |
+
</tr>
|
| 106 |
+
<tr>
|
| 107 |
+
<td>
|
| 108 |
+
<audio controls>
|
| 109 |
+
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 3 - Test 1 - Titan.wav?download=true" type="audio/wav">
|
| 110 |
+
Your browser does not support the audio element.
|
| 111 |
+
</audio>
|
| 112 |
+
</td>
|
| 113 |
+
<td>
|
| 114 |
+
<audio controls>
|
| 115 |
+
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 3 - Test 1 - Ov2.wav?download=true" type="audio/wav">
|
| 116 |
+
Your browser does not support the audio element.
|
| 117 |
+
</audio>
|
| 118 |
+
</td>
|
| 119 |
+
<td>
|
| 120 |
+
<audio controls>
|
| 121 |
+
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 3 - Test 1 - Ov2.1.wav?download=true" type="audio/wav">
|
| 122 |
+
Your browser does not support the audio element.
|
| 123 |
+
</audio>
|
| 124 |
+
</td>
|
| 125 |
+
</tr>
|
| 126 |
+
|
| 127 |
+
</tr>
|
| 128 |
+
<tr>
|
| 129 |
+
<td>
|
| 130 |
+
<audio controls>
|
| 131 |
+
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 3 - Test 2 - Titan.wav?download=true" type="audio/wav">
|
| 132 |
+
Your browser does not support the audio element.
|
| 133 |
+
</audio>
|
| 134 |
+
</td>
|
| 135 |
+
<td>
|
| 136 |
+
<audio controls>
|
| 137 |
+
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 3 - Test 2 - Ov2.wav?download=true" type="audio/wav">
|
| 138 |
+
Your browser does not support the audio element.
|
| 139 |
+
</audio>
|
| 140 |
+
</td>
|
| 141 |
+
<td>
|
| 142 |
+
<audio controls>
|
| 143 |
+
<source src="https://huggingface.co/blaise-tk/TITAN/resolve/main/demos/Model 3 - Test 2 - Ov2.1.wav?download=true" type="audio/wav">
|
| 144 |
+
Your browser does not support the audio element.
|
| 145 |
+
</audio>
|
| 146 |
+
</td>
|
| 147 |
+
</tr>
|
| 148 |
+
|
| 149 |
+
</table>
|
| 150 |
+
|
| 151 |
+
### Titan-Large
|
| 152 |
+
|
| 153 |
+
- Details forthcoming...
|
| 154 |
+
|
| 155 |
+
## Collaborators
|
| 156 |
+
|
| 157 |
+
We appreciate the contributions of our collaborators who have helped in the development and refinement of TITAN.
|
| 158 |
+
|
| 159 |
+
- Mustar
|
| 160 |
+
- SimplCup
|
| 161 |
+
- UnitedShoes
|
| 162 |
+
|
| 163 |
+
## Beta Testers
|
| 164 |
+
|
| 165 |
+
We extend our gratitude to the beta testers who provided valuable feedback during the testing phase of TITAN.
|
| 166 |
+
|
| 167 |
+
- SimplCup
|
| 168 |
+
- Leo_Frixi
|
| 169 |
+
- Light
|
| 170 |
+
- SCRFilms
|
| 171 |
+
- Ryanz
|
| 172 |
+
- Litsa_the_dancer
|
| 173 |
+
|
| 174 |
+
## Citation
|
| 175 |
+
|
| 176 |
+
Should you find TITAN beneficial for your research endeavors or projects, we kindly request citing our repository:
|
| 177 |
+
|
| 178 |
+
```
|
| 179 |
+
@article{titan,
|
| 180 |
+
title={TITAN: A Versatile, Robust, and High-Quality Pretrained Model for Retrieval-based Voice Conversion (RVC) Training},
|
| 181 |
+
author={Blaise},
|
| 182 |
+
journal={Hugging Face},
|
| 183 |
+
year={2024},
|
| 184 |
+
publisher={Blaise},
|
| 185 |
+
url={https://huggingface.co/blaise-tk/TITAN/}
|
| 186 |
+
}
|
| 187 |
+
```
|