Galuh
commited on
Commit
·
e74c80d
1
Parent(s):
fab5852
Update README.md
Browse files
README.md
CHANGED
@@ -86,7 +86,7 @@ processor = Wav2Vec2Processor.from_pretrained("indonesian-nlp/wav2vec2-large-xls
|
|
86 |
model = Wav2Vec2ForCTC.from_pretrained("indonesian-nlp/wav2vec2-large-xlsr-indonesian-baseline")
|
87 |
model.to("cuda")
|
88 |
|
89 |
-
chars_to_ignore_regex = '[
|
90 |
|
91 |
resampler = torchaudio.transforms.Resample(48_000, 16_000)
|
92 |
|
@@ -123,5 +123,4 @@ print("WER: {:2f}".format(100 * wer.compute(predictions=result["pred_strings"],
|
|
123 |
|
124 |
The Common Voice `train`, `validation`, and ... datasets were used for training as well as ... and ... # TODO
|
125 |
|
126 |
-
The script used for training can be found [here](https://github.com/
|
127 |
-
(will be available soon)
|
|
|
86 |
model = Wav2Vec2ForCTC.from_pretrained("indonesian-nlp/wav2vec2-large-xlsr-indonesian-baseline")
|
87 |
model.to("cuda")
|
88 |
|
89 |
+
chars_to_ignore_regex = '[\\\\\\\\,\\\\\\\\?\\\\\\\\.\\\\\\\\!\\\\\\\\-\\\\\\\\;\\\\\\\\:\\\\\\\\"\\\\\\\\“]'
|
90 |
|
91 |
resampler = torchaudio.transforms.Resample(48_000, 16_000)
|
92 |
|
|
|
123 |
|
124 |
The Common Voice `train`, `validation`, and ... datasets were used for training as well as ... and ... # TODO
|
125 |
|
126 |
+
The script used for training can be found [here](https://github.com/indonesian-nlp/wav2vec2-indonesian)
|
|