Fix double backslashs
Browse files
README.md
CHANGED
|
@@ -39,7 +39,7 @@ model = Wav2Vec2ForCTC.from_pretrained("Marxav/wav2vec2-large-xlsr-53-breton")
|
|
| 39 |
|
| 40 |
resampler = torchaudio.transforms.Resample(48_000, 16_000)
|
| 41 |
|
| 42 |
-
chars_to_ignore_regex = '[
|
| 43 |
|
| 44 |
# Preprocessing the datasets.
|
| 45 |
# We need to read the audio files as arrays
|
|
@@ -67,7 +67,7 @@ print("Reference:", test_dataset["sentence"][:nb_samples])
|
|
| 67 |
```
|
| 68 |
The above code leads to the following prediction for the first two samples:
|
| 69 |
* Prediction: ["nel ler ket dont abenn eus netra la vez ser mirc'hid evel sij", 'an eil hag egile']
|
| 70 |
-
* Reference: [
|
| 71 |
|
| 72 |
The model can be evaluated as follows on the {language} test data of Common Voice.
|
| 73 |
```python
|
|
@@ -85,7 +85,7 @@ processor = Wav2Vec2Processor.from_pretrained('Marxav/wav2vec2-large-xlsr-53-bre
|
|
| 85 |
model = Wav2Vec2ForCTC.from_pretrained('Marxav/wav2vec2-large-xlsr-53-breton2')
|
| 86 |
model.to("cuda")
|
| 87 |
|
| 88 |
-
chars_to_ignore_regex = '[
|
| 89 |
|
| 90 |
resampler = torchaudio.transforms.Resample(48_000, 16_000)
|
| 91 |
|
|
|
|
| 39 |
|
| 40 |
resampler = torchaudio.transforms.Resample(48_000, 16_000)
|
| 41 |
|
| 42 |
+
chars_to_ignore_regex = '[\\,\,\?\.\!\;\:\"\“\%\”\�\(\)\/\«\»\½\…]'
|
| 43 |
|
| 44 |
# Preprocessing the datasets.
|
| 45 |
# We need to read the audio files as arrays
|
|
|
|
| 67 |
```
|
| 68 |
The above code leads to the following prediction for the first two samples:
|
| 69 |
* Prediction: ["nel ler ket dont abenn eus netra la vez ser mirc'hid evel sij", 'an eil hag egile']
|
| 70 |
+
* Reference: ["N'haller ket dont a-benn eus netra pa vezer nec'het evel-se ", 'An eil hag egile ']
|
| 71 |
|
| 72 |
The model can be evaluated as follows on the {language} test data of Common Voice.
|
| 73 |
```python
|
|
|
|
| 85 |
model = Wav2Vec2ForCTC.from_pretrained('Marxav/wav2vec2-large-xlsr-53-breton2')
|
| 86 |
model.to("cuda")
|
| 87 |
|
| 88 |
+
chars_to_ignore_regex = '[\\,\,\?\.\!\;\:\"\“\%\”\�\(\)\/\«\»\½\…]'
|
| 89 |
|
| 90 |
resampler = torchaudio.transforms.Resample(48_000, 16_000)
|
| 91 |
|