Update app.py

app.py

@@ -139,7 +139,7 @@ def generate_answer(text_input):
         return f"Error: {str(e)}"
 
 def process_audio_stream(audio_path, question_text=None):
-    """Streaming pipeline for Gradio UI"""
+    """Streaming pipeline for Gradio UI - Returns tuple generator"""
     start_time = time.time()
     
     # Step 1: Transcribe audio if provided
@@ -170,6 +170,7 @@ def process_audio_stream(audio_path, question_text=None):
         current_time = time.time() - start_time
         time_emoji = "🟢" if current_time < 3.0 else "🟡" if current_time < 3.5 else "🔴"
         timing_info = f"\n\n{time_emoji} **Timing:** Trans={transcription_time:.2f}s | Search={search_time:.2f}s | LLM={(time.time()-llm_start):.2f}s | **Total={current_time:.2f}s**"
+        # IMPORTANT: Must yield tuple (text, number) to match output components
         yield partial_answer + timing_info, current_time
 
 # Create Gradio interface
@@ -195,9 +196,10 @@ with gr.Blocks(title="Fast Q&A - Streaming Enabled", theme=gr.themes.Soft()) as
                 audio_output = gr.Textbox(label="Answer (Streaming)", lines=8, show_copy_button=True)
                 audio_time = gr.Number(label="Response Time (seconds)", precision=2)
         
+        # Fixed: Lambda wrapper ensures proper tuple unpacking
         audio_submit.click(
-            fn=lambda x: process_audio_stream(x, None),
-            inputs=[audio_input],
+            fn=process_audio_stream,
+            inputs=[audio_input, gr.Textbox(value=None, visible=False)],
             outputs=[audio_output, audio_time],
             api_name="audio_query_stream"
         )
@@ -216,8 +218,9 @@ with gr.Blocks(title="Fast Q&A - Streaming Enabled", theme=gr.themes.Soft()) as
                 text_output = gr.Textbox(label="Answer (Streaming)", lines=8, show_copy_button=True)
                 text_time = gr.Number(label="Response Time (seconds)", precision=2)
         
+        # Fixed: Proper function call with audio=None
         text_submit.click(
-            fn=lambda x: process_audio_stream(None, x),
+            fn=lambda text: process_audio_stream(None, text),
             inputs=[text_input],
             outputs=[text_output, text_time],
             api_name="text_query_stream"
@@ -232,7 +235,7 @@ with gr.Blocks(title="Fast Q&A - Streaming Enabled", theme=gr.themes.Soft()) as
             inputs=text_input
         )
     
-    # API endpoints for Pluely - Using Gradio's built-in interface functions
+    # API endpoints for Pluely
     with gr.Tab("🔌 Pluely Integration"):
         gr.Markdown("""
         ## Dedicated Endpoints for Pluely
@@ -278,8 +281,8 @@ with gr.Blocks(title="Fast Q&A - Streaming Enabled", theme=gr.themes.Soft()) as
         with gr.Row(visible=False):
             stt_input = gr.Textbox()
             stt_output = gr.JSON()
-            ai_input = gr.Textbox()
-            ai_output = gr.Textbox()
+            ai_stream_input = gr.Textbox()
+            ai_stream_output = gr.Textbox()
         
         # These create the /call/transcribe_stt and /call/answer_ai_stream endpoints
         stt_button = gr.Button("STT", visible=False)
@@ -290,11 +293,11 @@ with gr.Blocks(title="Fast Q&A - Streaming Enabled", theme=gr.themes.Soft()) as
             api_name="transcribe_stt"
         )
         
-        ai_button = gr.Button("AI", visible=False)
-        ai_button.click(
+        ai_stream_button = gr.Button("AI Stream", visible=False)
+        ai_stream_button.click(
             fn=generate_answer_stream,
-            inputs=[ai_input],
-            outputs=[ai_output],
+            inputs=[ai_stream_input],
+            outputs=[ai_stream_output],
             api_name="answer_ai_stream"
         )