Update audio2text/a2t.py
Browse files- audio2text/a2t.py +3 -4
audio2text/a2t.py
CHANGED
@@ -40,15 +40,14 @@ class A2T:
|
|
40 |
|
41 |
def predict(self):
|
42 |
if self.mic is not None:
|
43 |
-
audio = self.mic
|
44 |
-
|
45 |
-
frame_rate = self.mic.frame_rate
|
46 |
else:
|
47 |
return "please provide audio"
|
48 |
|
49 |
try:
|
50 |
forced_decoder_ids = processor.get_decoder_prompt_ids(language="english", task="transcribe")
|
51 |
-
audio = self.__preprocces(audio=audio, frame_rate=frame_rate)
|
52 |
inputs = processor(audio=audio, sampling_rate=16000, return_tensors="pt")
|
53 |
predicted_ids = model.generate(**inputs, max_length=400, forced_decoder_ids=forced_decoder_ids)
|
54 |
transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
|
|
|
40 |
|
41 |
def predict(self):
|
42 |
if self.mic is not None:
|
43 |
+
audio = self.mic
|
44 |
+
# frame_rate = self.mic.frame_rate
|
|
|
45 |
else:
|
46 |
return "please provide audio"
|
47 |
|
48 |
try:
|
49 |
forced_decoder_ids = processor.get_decoder_prompt_ids(language="english", task="transcribe")
|
50 |
+
# audio = self.__preprocces(audio=audio, frame_rate=frame_rate)
|
51 |
inputs = processor(audio=audio, sampling_rate=16000, return_tensors="pt")
|
52 |
predicted_ids = model.generate(**inputs, max_length=400, forced_decoder_ids=forced_decoder_ids)
|
53 |
transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
|