Update app.py

app.py

@@ -13,7 +13,7 @@ import subprocess
 
 default_lang = "en"
 
-engines = {default_lang: Model(default_lang)}
+engines = { default_lang: Model(default_lang) }
 
 def transcribe(audio):
     if audio is None:
@@ -175,6 +175,16 @@ def voice_assistant_tab():
 def speech_translation_tab():
     return "# <center><b>Hear how you sound in another language</b></center>"
 
+# New function to handle streaming audio input
+def process_audio(audio, model, seed):
+    if audio is None:
+        return None
+    user_input = transcribe(audio)
+    if not user_input:
+        return None
+    reply = models(user_input, model, seed)
+    return asyncio.run(respond(audio, model, seed))
+
 with gr.Blocks(css="style.css") as demo:
     description = gr.Markdown("# <center><b>Hello, I am Optimus Prime your personal AI voice assistant</b></center>")
     
@@ -198,21 +208,29 @@ with gr.Blocks(css="style.css") as demo:
             value=0,
             visible=False
             )
-            input = gr.Audio(label="User", sources="microphone", type="filepath", live=True)
-            output = gr.Audio(label="AI", type="filepath",
-                            interactive=False,
-                            autoplay=True,
-                            elem_classes="audio")
+            input = gr.Audio(
+                label="User",
+                sources="microphone",
+                type="filepath",
+                streaming=True,
+                waveform_options={"show_controls": False, "height": 100}
+            )
+            output = gr.Audio(
+                label="AI",
+                type="filepath",
+                interactive=False,
+                autoplay=True,
+                elem_classes="audio"
+            )
 
-            gr.Interface(
-                fn=respond, 
+            input.stream(
+                fn=process_audio,
                 inputs=[input, select, seed],
-                outputs=[output],
-                live=True
+                outputs=[output]
             )
         
         with gr.TabItem("Speech Translation") as speech_translation:
-            input_audio = gr.Audio(label="User", sources="microphone", type="filepath", live=True)
+            input_audio = gr.Audio(label="User", sources="microphone", type="filepath", waveform_options=False)
             target_lang = gr.Dropdown(
                 choices=list(LANGUAGE_CODES.keys()),
                 value="German",
@@ -230,16 +248,8 @@ with gr.Blocks(css="style.css") as demo:
                 live=True
             )
 
-    # clear_button = gr.Button("Clear")
-    # clear_button.click(
-    #     fn=clear_history,
-    #     inputs=[],
-    #     outputs=[input, output, input_audio, output_audio],
-    #     api_name="clear"
-    # )
-
     voice_assistant.select(fn=voice_assistant_tab, inputs=None, outputs=description)
     speech_translation.select(fn=speech_translation_tab, inputs=None, outputs=description)
 
 if __name__ == "__main__":
-    demo.queue(max_size=200).launch()
+    demo.queue(max_size=200).launch()