Update app.py

app.py

@@ -19,39 +19,20 @@
 
 from speechbrain.inference.ASR import EncoderASR
 import gradio as gr
-import numpy as np
-import sounddevice as sd
-import soundfile as sf
 
 model = EncoderASR.from_hparams("speechbrain/asr-wav2vec2-dvoice-wolof")
 
-def transcribe_audio(audio):
-    return model.transcribe_file(audio.name)
-
-def transcribe_microphone():
-    # Record audio from microphone for 5 seconds
-    duration = 60  # seconds
-    fs = 16000  # sampling rate
-    recording = sd.rec(int(duration * fs), samplerate=fs, channels=1, dtype=np.int16)
-    sd.wait()
-
-    # Save recorded audio to a temporary file
-    temp_audio_file = "temp_audio.wav"
-    sf.write(temp_audio_file, recording, fs)
-
-    # Transcribe the recorded audio
-    transcription = model.transcribe_file(temp_audio_file)
-
-    return transcription
+def transcribe(audio_or_file):
+    if isinstance(audio_or_file, str):  # If input is a file path
+        return model.transcribe_file(audio_or_file)
+    else:  # If input is audio from microphone
+        return model.transcribe_array(audio_or_file)
 
 demo = gr.Interface(
-    fn=[transcribe_audio, transcribe_microphone],
-    inputs=["file", "microphone"],
-    outputs="text",
-    title="Transcription automatique du wolof",
-    description="Ce modèle transcrit un fichier audio en wolof en texte en utilisant l'alphabet latin.",
-    input_labels=["Audio en wolof", "Microphone (parlez en wolof)"],
-    output_label="Transcription alphabet latin"
+    transcribe,
+    ["state", gr.inputs.Microphone(source="local", type="wav"), "text"],
+    ["state", "text"],
+    live=True,
 )
 
 demo.launch()