Update README.md
Browse files
README.md
CHANGED
|
@@ -3,9 +3,11 @@ license: mit
|
|
| 3 |
---
|
| 4 |
|
| 5 |
# hubert-base-jtube
|
| 6 |
-
This repo provides model weights for the [hubert-base model](https://arxiv.org/abs/2106.07447) trained on the [JTubeSpeech](https://github.com/sarulab-speech/jtubespeech) corpus
|
| 7 |
-
## Dataset
|
| 8 |
|
|
|
|
|
|
|
|
|
|
| 9 |
|
| 10 |
# Contributors
|
| 11 |
* [中田 亘](https://wataru-nakata.github.io)
|
|
|
|
| 3 |
---
|
| 4 |
|
| 5 |
# hubert-base-jtube
|
| 6 |
+
This repo provides model weights for the [hubert-base model](https://arxiv.org/abs/2106.07447) trained on the [JTubeSpeech](https://github.com/sarulab-speech/jtubespeech) corpus.
|
|
|
|
| 7 |
|
| 8 |
+
## Dataset
|
| 9 |
+
We extracted approximately 2720 hours of Japanese speech from the single-speaker subset of the JTubeSpeech corpus.
|
| 10 |
+
The training data includes approximately 6,000,000 utterances from a total of about 55,000 speakers.
|
| 11 |
|
| 12 |
# Contributors
|
| 13 |
* [中田 亘](https://wataru-nakata.github.io)
|