Spaces:
Build error
Build error
Update app.py
Browse files
app.py
CHANGED
|
@@ -5,21 +5,15 @@ import whisper
|
|
| 5 |
model = whisper.load_model("base")
|
| 6 |
|
| 7 |
def transcribe(audio_file):
|
| 8 |
-
# Process the audio file
|
| 9 |
-
|
| 10 |
-
audio = whisper.pad_or_trim(audio)
|
| 11 |
-
|
| 12 |
-
# Make predictions
|
| 13 |
-
mel = whisper.log_mel_spectrogram(audio).to(model.device)
|
| 14 |
-
options = whisper.DecodingOptions(fp16=False)
|
| 15 |
-
result = whisper.decode(model, mel, options)
|
| 16 |
|
| 17 |
# Return the transcription
|
| 18 |
-
return result
|
| 19 |
|
| 20 |
# Create the Gradio interface
|
| 21 |
iface = gr.Interface(fn=transcribe,
|
| 22 |
-
inputs=gr.Audio(sources="upload", type="
|
| 23 |
outputs="text",
|
| 24 |
title="Whisper Transcription",
|
| 25 |
description="Upload an audio file to transcribe it using OpenAI's Whisper model.")
|
|
|
|
| 5 |
model = whisper.load_model("base")
|
| 6 |
|
| 7 |
def transcribe(audio_file):
|
| 8 |
+
# Process the audio file directly with the file path
|
| 9 |
+
result = model.transcribe(audio_file)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
|
| 11 |
# Return the transcription
|
| 12 |
+
return result['text']
|
| 13 |
|
| 14 |
# Create the Gradio interface
|
| 15 |
iface = gr.Interface(fn=transcribe,
|
| 16 |
+
inputs=gr.inputs.Audio(sources="upload", type="file", label="Upload Audio"),
|
| 17 |
outputs="text",
|
| 18 |
title="Whisper Transcription",
|
| 19 |
description="Upload an audio file to transcribe it using OpenAI's Whisper model.")
|