Update README.md
Browse files
README.md
CHANGED
@@ -21,9 +21,20 @@ Our preprocessing pipeline includes adaptive context-aware noise injection to en
|
|
21 |
##### Example Usage
|
22 |
|
23 |
```python
|
|
|
|
|
|
|
24 |
data_collator = DataCollatorSpeechSeq2SeqWithPadding(
|
25 |
processor=processor,
|
26 |
decoder_start_token_id=model.config.decoder_start_token_id,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
27 |
apply_augmentation=True,
|
28 |
apply_noise_injection=True # Enable adaptive noise injection
|
29 |
)
|
|
|
21 |
##### Example Usage
|
22 |
|
23 |
```python
|
24 |
+
|
25 |
+
## HF transformers
|
26 |
+
|
27 |
data_collator = DataCollatorSpeechSeq2SeqWithPadding(
|
28 |
processor=processor,
|
29 |
decoder_start_token_id=model.config.decoder_start_token_id,
|
30 |
+
apply_augmentation=True, # Enable augmentation
|
31 |
+
apply_noise_injection=True # Enable adaptive noise injection
|
32 |
+
)
|
33 |
+
|
34 |
+
|
35 |
+
### pytorch
|
36 |
+
|
37 |
+
data_collator = DataCollatorSpeechSeq2SeqWithPadding(
|
38 |
apply_augmentation=True,
|
39 |
apply_noise_injection=True # Enable adaptive noise injection
|
40 |
)
|