Update app.py

app.py

@@ -52,7 +52,7 @@ def process_audio(audio_data):
 # 定义函数以禁用按钮并显示加载指示器
 def disable_components():
     # 更新 recognized_text 的内容，提示用户正在处理
-    recognized_text_update = gr.update(value='正在处理，请稍候...')
+    recognized_text_update = gr.update(value='Voice Recognization Running...')
     # 禁用 process_button
     process_button_update = gr.update(interactive=False)
     # 显示加载动画
@@ -67,12 +67,14 @@ def enable_components(recognized_text):
     return recognized_text, process_button_update, loading_animation_update
 
 llama_responded = 0
+responded_answer = ""
 
 def respond(
     message,
     history: list[tuple[str, str]]
 ):
     global llama_responded
+    global responded_answer
     system_message = "You are a helpful chatbot that answers questions. Give any answer within 50 words."
     messages = [{"role": "system", "content": system_message}]
 
@@ -95,42 +97,52 @@ def respond(
         response += token
 
     llama_responded = 1
+    responded_answer = response
     return response
 
 def update_response_display():
     while not llama_responded:
         time.sleep(1)
 
-def bot(history):
+def tts_part():
     global llama_responded
-    #print(history)
-    history.append([None,gr.Audio("01.wav")])
-    llama_responded = 0
-   
-    return history
+    global responded_answer
+    result = ""
+    if responded_answer != "":
+        text = responded_answer
+
+        client = Client("tonyassi/voice-clone")
+        result = client.predict(
+                text,
+                audio=file('siri.wav'),
+                api_name="/predict"
+        )
+        llama_responded = 0
+        responded_answer = ""
+    return result
 
 def create_interface():
     with gr.Blocks() as demo:
-        
+
         # Chat interface using the custom chatbot instance
         chatbot = gr.ChatInterface(
             title="Exodia AI Assistant",
+            fill_height=True,
             fn=respond,
             submit_btn="Start Chatting"
         )
-        
         user_start = chatbot.textbox.submit(
             fn=update_response_display, 
             inputs=[],
             outputs=[],
         )
-
-        user_start.then(
-            fn=bot,
-            inputs=chatbot.chatbot, 
-            outputs=chatbot.chatbot
+        user_click = chatbot.submit_btn.click(
+            fn=update_response_display, 
+            inputs=[],
+            outputs=[],
         )
 
+
         # Audio input section
         with gr.Row():
             audio_input = gr.Audio(
@@ -148,6 +160,10 @@ def create_interface():
             value='<div style="text-align: center;"><span style="font-size: 18px;">ASR Model is running...</span></div>',
             visible=False
         )
+
+        text_speaker = gr.Audio(
+            label="Generated Audio"
+        )
         
         # Associate audio processing function and update component states on click
         process_button.click(
@@ -163,12 +179,23 @@ def create_interface():
             inputs=[recognized_text],
             outputs=[recognized_text, process_button, loading_animation]
         )
+
+        user_start.then(
+            fn=tts_part,
+            inputs=[], 
+            outputs=text_speaker
+        )
+
+        user_click.then(
+            fn=tts_part,
+            inputs=[], 
+            outputs=text_speaker
+        )
                 
     return demo
 
 
 
-
 if __name__ == "__main__":
     demo = create_interface()
     demo.launch()