Update README.md
Browse files
README.md
CHANGED
|
@@ -21,3 +21,7 @@ configs:
|
|
| 21 |
- split: train
|
| 22 |
path: data/train-*
|
| 23 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
- split: train
|
| 22 |
path: data/train-*
|
| 23 |
---
|
| 24 |
+
|
| 25 |
+
Synthetic dataset based on highly specialised texts from Wikipedia. Voiced using Yandex SpeechKit with random voices, roles and speech rate.
|
| 26 |
+
Can be used to evaluate ASR models not trained on given domains and to identify areas that the model does not handle well.
|
| 27 |
+
Splitted by category version can be found here: https://huggingface.co/datasets/rmndrnts/wikipedia_asr_splitted
|