Update app.py
Browse files
app.py
CHANGED
|
@@ -19,14 +19,6 @@ torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
|
|
| 19 |
|
| 20 |
print(f"Using device: {device}")
|
| 21 |
|
| 22 |
-
# Initialize the ASR pipeline from Hugging Face Transformers
|
| 23 |
-
transcriber = pipeline(
|
| 24 |
-
"automatic-speech-recognition",
|
| 25 |
-
model="openai/whisper-large-v3-turbo",
|
| 26 |
-
torch_dtype=torch_dtype,
|
| 27 |
-
device=device,
|
| 28 |
-
)
|
| 29 |
-
|
| 30 |
# --- Transcription Function ---
|
| 31 |
def extract_audio_from_video(video_path, output_audio_path="temp_extracted_audio.wav"):
|
| 32 |
"""
|
|
@@ -60,6 +52,16 @@ def transcribe_audio(audio_path):
|
|
| 60 |
This function takes an audio file path, transcribes it using the Whisper model,
|
| 61 |
and returns the transcribed text.
|
| 62 |
"""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 63 |
if audio_path is None:
|
| 64 |
return "No audio file provided. Please upload or record an audio file."
|
| 65 |
|
|
|
|
| 19 |
|
| 20 |
print(f"Using device: {device}")
|
| 21 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
# --- Transcription Function ---
|
| 23 |
def extract_audio_from_video(video_path, output_audio_path="temp_extracted_audio.wav"):
|
| 24 |
"""
|
|
|
|
| 52 |
This function takes an audio file path, transcribes it using the Whisper model,
|
| 53 |
and returns the transcribed text.
|
| 54 |
"""
|
| 55 |
+
# Initialize the ASR pipeline from Hugging Face Transformers
|
| 56 |
+
|
| 57 |
+
transcriber = pipeline(
|
| 58 |
+
"automatic-speech-recognition",
|
| 59 |
+
model="openai/whisper-large-v3-turbo",
|
| 60 |
+
torch_dtype=torch_dtype,
|
| 61 |
+
device=device,
|
| 62 |
+
language='en',
|
| 63 |
+
)
|
| 64 |
+
|
| 65 |
if audio_path is None:
|
| 66 |
return "No audio file provided. Please upload or record an audio file."
|
| 67 |
|