Update app.py

app.py

@@ -701,10 +701,9 @@ with gr.Blocks() as iface:
     gr.Markdown("Transcribe audio using multiple pipelines and (Faster) Whisper models.")
     
     with gr.Row():
-        audio_upload = gr.Audio(label="Upload Audio")
-        audio_record = gr.Audio(label="Record Audio", source="microphone")
+        audio_input = gr.Audio(label="Upload or Record Audio", sources=["upload", "microphone"], type="filepath")
         audio_url = gr.Textbox(label="Or Enter URL of audio file or YouTube link")
-    
+
     transcribe_button = gr.Button("Transcribe")
 
     with gr.Accordion("Advanced Options", open=False):
@@ -771,15 +770,13 @@ with gr.Blocks() as iface:
     pipeline_type.change(update_model_dropdown, inputs=[pipeline_type], outputs=[model_id])
     
     def transcribe_with_progress(*args):
-        # Combine audio_upload and audio_record into a single input
-        audio_input = args[0] if args[0] is not None else args[1]
-        modified_args = (audio_input,) + args[2:]
-        for result in transcribe_audio(*modified_args):
+        # The audio_input is now the first argument
+        for result in transcribe_audio(*args):
             yield result
     
     transcribe_button.click(
         transcribe_with_progress,
-        inputs=[audio_upload, audio_record, audio_url, proxy_url, proxy_username, proxy_password, pipeline_type, model_id, dtype, batch_size, download_method, start_time, end_time, verbose, include_timecodes],
+        inputs=[audio_input, audio_url, proxy_url, proxy_username, proxy_password, pipeline_type, model_id, dtype, batch_size, download_method, start_time, end_time, verbose, include_timecodes],
         outputs=[metrics_output, transcription_output, transcription_file]
     )