Update app.py

app.py

@@ -34,8 +34,7 @@ def parse_transcription(logits):
     transcription = processor.decode(predicted_ids[0], skip_special_tokens=True)
     return transcription
 
-def transcribe(audio, audio_microphone, applyLM):
-    audio_path = audio_microphone if audio_microphone else audio
+def transcribe(audio_path, applyLM):
     speech_array, sampling_rate = torchaudio.load(audio_path)
     speech = torchaudio.functional.resample(speech_array, orig_freq=sampling_rate, new_freq=16000).squeeze().numpy()
     """
@@ -64,8 +63,7 @@ gradio_ui = gr.Interface(
     fn=transcribe,
     title="Kikuyu Speech Recognition",
     description="",
-    inputs=[gr.Audio(label="Upload Audio File", type="filepath", optional=True), 
-            gr.Audio(source="microphone", type="filepath", optional=True, label="Record from microphone"),
+    inputs=[gr.Audio(source="microphone", type="filepath", optional=True, label="Record from microphone"),
             gr.Checkbox(label="Apply LM", value=False)],
     outputs=[gr.outputs.Textbox(label="Recognized speech")]
 )