Update README.md
Browse files
README.md
CHANGED
|
@@ -43,8 +43,8 @@ from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
|
|
| 43 |
|
| 44 |
test_dataset = load_dataset("common_voice", "tr", split="test[:2%]")
|
| 45 |
|
| 46 |
-
processor = Wav2Vec2Processor.from_pretrained("ceyda/wav2vec2-large-xlsr-turkish")
|
| 47 |
-
model = Wav2Vec2ForCTC.from_pretrained("ceyda/wav2vec2-large-xlsr-turkish")
|
| 48 |
|
| 49 |
resampler = torchaudio.transforms.Resample(48_000, 16_000)
|
| 50 |
|
|
@@ -86,8 +86,8 @@ processor = Wav2Vec2Processor.from_pretrained("ceyda/wav2vec2-large-xlsr-turkish
|
|
| 86 |
model = Wav2Vec2ForCTC.from_pretrained("ceyda/wav2vec2-large-xlsr-turkish")
|
| 87 |
model.to("cuda")
|
| 88 |
|
| 89 |
-
# chars_to_ignore_regex = '[
|
| 90 |
-
chars_to_ignore_regex = '[
|
| 91 |
resampler = torchaudio.transforms.Resample(48_000, 16_000)
|
| 92 |
|
| 93 |
# Preprocessing the datasets.
|
|
|
|
| 43 |
|
| 44 |
test_dataset = load_dataset("common_voice", "tr", split="test[:2%]")
|
| 45 |
|
| 46 |
+
processor = Wav2Vec2Processor.from_pretrained("ceyda/wav2vec2-large-xlsr-53-turkish")
|
| 47 |
+
model = Wav2Vec2ForCTC.from_pretrained("ceyda/wav2vec2-large-xlsr-53-turkish")
|
| 48 |
|
| 49 |
resampler = torchaudio.transforms.Resample(48_000, 16_000)
|
| 50 |
|
|
|
|
| 86 |
model = Wav2Vec2ForCTC.from_pretrained("ceyda/wav2vec2-large-xlsr-turkish")
|
| 87 |
model.to("cuda")
|
| 88 |
|
| 89 |
+
# chars_to_ignore_regex = '[\\,\\?\\.\\!\\-\\;\\:\\"\\“]'
|
| 90 |
+
chars_to_ignore_regex = '[\\,\\?\\.\\!\\-\\;\\:\\"\\“\\%\\‘\\”\\�\\'\\`]'
|
| 91 |
resampler = torchaudio.transforms.Resample(48_000, 16_000)
|
| 92 |
|
| 93 |
# Preprocessing the datasets.
|