Update app.py

app.py

@@ -116,7 +116,7 @@ def generate_response(user_input):
     with torch.no_grad():
         outputs = proc_model.generate(
             inputs, 
-            max_new_tokens=256,
+            max_new_tokens=100,
         )
     
     response = proc_tokenizer.batch_decode(
@@ -129,14 +129,14 @@ def generate_response(user_input):
     
 def CanaryPhiVits(user_voice):
     user_input = transcribe(user_voice)
-    print(user_input)
+    print("user_input:")
+    print(user_input)    
     response = generate_response(user_input)
-    print(response)
     if response.startswith(user_input):
         response = response.replace(user_input, '', 1)
+    print("chatty_response:")
     print(response)
     chatty_response = text_to_speech(response)
-
     
     return chatty_response
 
@@ -144,10 +144,19 @@ def CanaryPhiVits(user_voice):
 # Create a Gradio interface
 iface = gr.Interface(
     fn=CanaryPhiVits, 
-    inputs=gr.Audio(sources=["microphone", "upload"], type="filepath", format="wav",), 
-    #inputs=gr.Textbox(lines=5, placeholder="Enter your text here..."), 
-    #outputs=gr.Textbox(),
-    outputs=gr.Audio("response.wav"),
+    title="Chatty Ashe",
+    theme="gstaff/xkcd",
+    
+    inputs=gr.Audio(
+        sources=["microphone", "upload"], 
+        label="Input Audio",
+        type="filepath", 
+        format="wav",
+    ), 
+    outputs=gr.Audio(
+        "response.wav", 
+        label="Output Audio"
+    ),
 )
 
 # Launch the interface