Update README.md
Browse files
README.md
CHANGED
|
@@ -47,10 +47,17 @@ More information needed
|
|
| 47 |
|
| 48 |
## Training and evaluation data
|
| 49 |
|
| 50 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 51 |
|
| 52 |
## Training procedure
|
| 53 |
|
|
|
|
|
|
|
|
|
|
| 54 |
### Training hyperparameters
|
| 55 |
|
| 56 |
The following hyperparameters were used during training:
|
|
|
|
| 47 |
|
| 48 |
## Training and evaluation data
|
| 49 |
|
| 50 |
+
Training:
|
| 51 |
+
- [mozilla-foundation/common_voice_11_0](https://huggingface.co/datasets/mozilla-foundation/common_voice_11_0) (train+validation)
|
| 52 |
+
|
| 53 |
+
Evaluation:
|
| 54 |
+
- [mozilla-foundation/common_voice_11_0](https://huggingface.co/datasets/mozilla-foundation/common_voice_11_0) (test)
|
| 55 |
|
| 56 |
## Training procedure
|
| 57 |
|
| 58 |
+
Datasets were augmented on-the-fly using [audiomentations](https://github.com/iver56/audiomentations) via PitchShift and TimeStretch transformations at `p=0.3`.
|
| 59 |
+
A space is added between each Chinese character, as demonstrated in the original paper. Effectively, WER == CER in this case.
|
| 60 |
+
|
| 61 |
### Training hyperparameters
|
| 62 |
|
| 63 |
The following hyperparameters were used during training:
|