Commit
•
bf091e3
1
Parent(s):
dc098aa
Update README.md
Browse files
README.md
CHANGED
|
@@ -83,7 +83,6 @@ input_features = processor(sample["array"], sampling_rate=sample["sampling_rate"
|
|
| 83 |
predicted_ids = model.generate(input_features.to(device), forced_decoder_ids=forced_decoder_ids)
|
| 84 |
|
| 85 |
# decode
|
| 86 |
-
transcription = processor.batch_decode(predicted_ids)
|
| 87 |
transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
|
| 88 |
|
| 89 |
print(transcription)
|
|
@@ -121,7 +120,6 @@ dataset = dataset.cast_column("audio", Audio(sampling_rate=16000))
|
|
| 121 |
#print(dataset)
|
| 122 |
|
| 123 |
def normalize(batch):
|
| 124 |
-
"""Normalizes GOLD"""
|
| 125 |
batch["gold_text"] = whisper_norm(batch['sentence'])
|
| 126 |
return batch
|
| 127 |
|
|
|
|
| 83 |
predicted_ids = model.generate(input_features.to(device), forced_decoder_ids=forced_decoder_ids)
|
| 84 |
|
| 85 |
# decode
|
|
|
|
| 86 |
transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
|
| 87 |
|
| 88 |
print(transcription)
|
|
|
|
| 120 |
#print(dataset)
|
| 121 |
|
| 122 |
def normalize(batch):
|
|
|
|
| 123 |
batch["gold_text"] = whisper_norm(batch['sentence'])
|
| 124 |
return batch
|
| 125 |
|