Update app.py

app.py

@@ -17,9 +17,6 @@ app = FastAPI()
 # Serve static files
 app.mount("/static", StaticFiles(directory="static"), name="static")
 
-# Initialize ggwave instance
-instance = ggwave.init()
-
 # Initialize Groq client
 client = Groq(api_key=os.environ.get("GROQ_API_KEY"))
 
@@ -29,7 +26,7 @@ class TextInput(BaseModel):
 @app.get("/")
 async def serve_homepage():
     """Serve the chat interface HTML."""
-    return FileResponse("static/index.html")
+    return FileResponse("static/conv.html")
 
 @app.get("/conv/")
 async def serve_convpage():
@@ -77,7 +74,9 @@ def text_to_speech(input_text: TextInput):
 @app.post("/chat/")
 async def chat_with_llm(file: UploadFile = File(...)):
     """Process input WAV, send text to LLM, and return generated response as WAV."""
-    global instance
+    # Initialize ggwave instance
+    instance = ggwave.init()
+    
     # Read the file content into memory without saving to disk
     file_content = await file.read()
     
@@ -128,7 +127,7 @@ async def chat_with_llm(file: UploadFile = File(...)):
                 wf.writeframes(waveform_int16.tobytes())  # Write waveform as bytes
             
             buffer.seek(0)
-            
+            ggwave.free(instance)
             return Response(
                 content=buffer.getvalue(), 
                 media_type="audio/wav",
@@ -140,6 +139,7 @@ async def chat_with_llm(file: UploadFile = File(...)):
             
         except Exception as e:
             print(f"Error processing audio: {str(e)}")
+            ggwave.free(instance)
             return Response(
                 content=f"Error processing audio: {str(e)}",
                 media_type="text/plain",
@@ -151,7 +151,10 @@ async def continuous_chat(
     chat_history: Optional[str] = Form(None)
 ):
     """Process input WAV with chat history, send text to LLM, and return response as WAV."""
-    global instance
+    # Initialize ggwave instance
+    instance = ggwave.init()
+    
+    
     
     # Parse chat history if provided
     messages = [{"role": "system", "content": "you are a helpful assistant. answer always in one sentence"}]
@@ -212,6 +215,7 @@ async def continuous_chat(
                 wf.writeframes(waveform_int16.tobytes())
             
             buffer.seek(0)
+            ggwave.free(instance)
             
             return Response(
                 content=buffer.getvalue(),
@@ -224,6 +228,7 @@ async def continuous_chat(
             
         except Exception as e:
             print(f"Error processing audio: {str(e)}")
+            ggwave.free(instance)
             return Response(
                 content=f"Error processing audio: {str(e)}",
                 media_type="text/plain",