input file handling edit

app.py

@@ -10,8 +10,11 @@ asr = pipeline(
 )
 pipeline1 = Pipeline.from_pretrained("pyannote/speaker-segmentation")
 
-def diarization(file_input,microphone_input,selection):
-    audio = file_input if str(selection) == "Upload" else Path(microphone_input)
+def diarization(file_input,mic_input,selection):
+    mic_path = None if mic_input is None else mic_input.name
+    audio = file_input if selection == "Upload" else mic_path
+    if audio is None:
+        return "Please check your inputs!", ""
 
     speaker_output = pipeline1(audio)
     text_output = asr(audio,return_timestamps="word")
@@ -37,13 +40,13 @@ def diarization(file_input,microphone_input,selection):
 title = "Speech Recognition with Speaker Diarization"
 description = "Speaker Diarization is the act of attributing parts of the audio recording to different speakers. This space aims to distinguish the speakers and apply speech-to-text from a given input audio file. Pre-trained models from Pyannote[1] for the Speaker Diarization and [2]."
 article = "<p style='text-align: center'><a href='https://github.com/pyannote/pyannote-audio' target='_blank'>[1] Pyannote - Speaker Diarization model</a></p>"
-inputs = [gr.inputs.Audio(source="upload", type="filepath", label="Upload your audio file here:"),
-        gr.inputs.Audio(source="microphone", type="filepath",label="Or use your Microphone:"),
-        gr.inputs.Radio(["Upload","Microphone"],type="value",label="Select which input:")]
+inputs = [gr.inputs.Audio(source="upload", type="filepath", label="Upload your audio file here:", optional=True),
+        gr.inputs.Audio(source="microphone", type="file",label="Or use your Microphone:", optional=True),
+        gr.inputs.Radio(["Upload","Microphone"], type="value", label="Select which input:")]
 outputs = [gr.outputs.Textbox(type="auto", label="Diarized Output"),
         gr.outputs.Textbox(type="auto",label="Full ASR Text for comparison")]
-examples = [["test_audio1.wav","test_audio1.wav","Upload"],
-            ["test_audio2.wav","test_audio2.wav","Upload"]]
+examples = [["test_audio1.wav",None,"Upload"],
+            ["test_audio2.wav",None,"Upload"]]
 
 app = gr.Interface(fn=diarization,
                 inputs=inputs,