Update README.md
Browse files
README.md
CHANGED
|
@@ -25,6 +25,7 @@ model-index:
|
|
| 25 |
|
| 26 |
Wav2Vec2-Large-XLSR-Welsh
|
| 27 |
Fine-tuned facebook/wav2vec2-large-xlsr-53 on the Welsh Common Voice dataset.
|
|
|
|
| 28 |
|
| 29 |
When using this model, make sure that your speech input is sampled at 16kHz.
|
| 30 |
Test Result: 29.4%
|
|
@@ -81,7 +82,7 @@ model = Wav2Vec2ForCTC.from_pretrained("Srulikbdd/Wav2vec2-large-xlsr-welsh")
|
|
| 81 |
|
| 82 |
model.to("cuda")
|
| 83 |
|
| 84 |
-
chars_to_ignore_regex = '[
|
| 85 |
|
| 86 |
resampler = torchaudio.transforms.Resample(48_000, 16_000)
|
| 87 |
|
|
|
|
| 25 |
|
| 26 |
Wav2Vec2-Large-XLSR-Welsh
|
| 27 |
Fine-tuned facebook/wav2vec2-large-xlsr-53 on the Welsh Common Voice dataset.
|
| 28 |
+
The data was augmented using standard augmentation approach.
|
| 29 |
|
| 30 |
When using this model, make sure that your speech input is sampled at 16kHz.
|
| 31 |
Test Result: 29.4%
|
|
|
|
| 82 |
|
| 83 |
model.to("cuda")
|
| 84 |
|
| 85 |
+
chars_to_ignore_regex = '[\\\\\\\\,\\\\\\\\?\\\\\\\\.\\\\\\\\!\\\\\\\\-\\\\\\\\u2013\\\\\\\\u2014\\\\\\\\;\\\\\\\\:\\\\\\\\"\\\\\\\\\\\\\\\\%\\\\\\\\\\\\\\\\\\\\\\\\]'
|
| 86 |
|
| 87 |
resampler = torchaudio.transforms.Resample(48_000, 16_000)
|
| 88 |
|