Update README.md
Browse files
README.md
CHANGED
|
@@ -5,27 +5,31 @@ task_categories:
|
|
| 5 |
language:
|
| 6 |
- ru
|
| 7 |
---
|
| 8 |
-
|
|
|
|
|
|
|
|
|
|
| 9 |
Unlike the original dataset, this dataset also contains transcriptions for voice lines.
|
| 10 |
Transcriptions include stresses for words, even for words with just a single vowel.
|
| 11 |
Additionally, there is a `metadata_source.csv` file, that contains voice lines text “as-is”.
|
| 12 |
|
| 13 |
-
Dataset info (see details in `stats.txt`):
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
|
|
|
|
| 17 |
|
| 18 |
In addition to the text, there are also non-speech sound tags in `metadata.csv`
|
| 19 |
(see `non_speech_tags.txt` for details).
|
| 20 |
How `metadata.csv` was created:
|
| 21 |
1. Russian voice lines texts were extracted from the game and matched with audio samples
|
| 22 |
-
from
|
| 23 |
-
2. Stresses for words were automatically added with
|
| 24 |
3. Non-speech sound tags were added from the original voice lines text.
|
| 25 |
|
| 26 |
`metadata.csv` structure:
|
| 27 |
path to audio | text | speaker id (see `speaker_ids.json` file for details)
|
| 28 |
|
| 29 |
This dataset was made possible by the tools and resources created by the following people:
|
| 30 |
-
Rootreck
|
| 31 |
-
Den4ikA
|
|
|
|
| 5 |
language:
|
| 6 |
- ru
|
| 7 |
---
|
| 8 |
+
# WitchSpeech: Russian voice lines Witcher 3 TTS dataset
|
| 9 |
+
|
| 10 |
+
This is a repack of the dataset
|
| 11 |
+
[so-vits-svc-4.0-ru-The_Witcher_3_Wild_Hunt](https://huggingface.co/Rootreck/so-vits-svc-4.0-ru-The_Witcher_3_Wild_Hunt) for TTS purposes.
|
| 12 |
Unlike the original dataset, this dataset also contains transcriptions for voice lines.
|
| 13 |
Transcriptions include stresses for words, even for words with just a single vowel.
|
| 14 |
Additionally, there is a `metadata_source.csv` file, that contains voice lines text “as-is”.
|
| 15 |
|
| 16 |
+
Dataset info (see details in `stats.txt`):
|
| 17 |
+
|
| 18 |
+
**Sample rate**: 48 000
|
| 19 |
+
**Total time**: 21.49
|
| 20 |
+
**Number of speakers**: 34
|
| 21 |
|
| 22 |
In addition to the text, there are also non-speech sound tags in `metadata.csv`
|
| 23 |
(see `non_speech_tags.txt` for details).
|
| 24 |
How `metadata.csv` was created:
|
| 25 |
1. Russian voice lines texts were extracted from the game and matched with audio samples
|
| 26 |
+
from [so-vits-svc-4.0-ru-The_Witcher_3_Wild_Hunt](https://huggingface.co/Rootreck/so-vits-svc-4.0-ru-The_Witcher_3_Wild_Hunt) dataset.
|
| 27 |
+
2. Stresses for words were automatically added with [RUAccent](https://github.com/Den4ikAI/ruaccent) library.
|
| 28 |
3. Non-speech sound tags were added from the original voice lines text.
|
| 29 |
|
| 30 |
`metadata.csv` structure:
|
| 31 |
path to audio | text | speaker id (see `speaker_ids.json` file for details)
|
| 32 |
|
| 33 |
This dataset was made possible by the tools and resources created by the following people:
|
| 34 |
+
[Rootreck](https://huggingface.co/Rootreck)
|
| 35 |
+
[Den4ikA](https://github.com/Den4ikAI)
|