Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -52,14 +52,6 @@ The french subset of the dataset [Multilingual TEDx](https://www.openslr.org/100
|
|
| 52 |
|
| 53 |
#### Configs:
|
| 54 |
- single_samples (default): all samples taken separately
|
| 55 |
-
- max=30s: combine consecutive samples for a period shorter than 30 seconds
|
| 56 |
-
- max=10s: combine consecutive samples for a period shorter than 10 seconds
|
| 57 |
-
- max: combine all the samples of a TEDx talk
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
#### dependencies (only needed for much faster audio decoding):
|
| 61 |
-
- ffmpeg: apt install ffmpeg
|
| 62 |
-
- ffmpeg-python: pip install ffmpeg-python
|
| 63 |
|
| 64 |
#### Sample
|
| 65 |
```
|
|
|
|
| 52 |
|
| 53 |
#### Configs:
|
| 54 |
- single_samples (default): all samples taken separately
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 55 |
|
| 56 |
#### Sample
|
| 57 |
```
|