made changes to punkt

app/main.py

@@ -439,10 +439,21 @@ async def startup_event():
     # Download NLTK data
     logger.info("Downloading NLTK data...")
     try:
-        await asyncio.to_thread(nltk.download, 'punkt', download_dir=nltk_data_dir, quiet=True)
-        await asyncio.to_thread(nltk.download, 'punkt_tab', download_dir=nltk_data_dir, quiet=True)
+        # Check if punkt is already downloaded
+        import nltk.data
+        try:
+            nltk.data.find('tokenizers/punkt', paths=[nltk_data_dir])
+            logger.info("NLTK punkt already downloaded")
+        except LookupError:
+            await asyncio.to_thread(nltk.download, 'punkt', download_dir=nltk_data_dir, quiet=True)
+            
+        try:
+            nltk.data.find('tokenizers/punkt_tab', paths=[nltk_data_dir])
+            logger.info("NLTK punkt_tab already downloaded")
+        except LookupError:
+            await asyncio.to_thread(nltk.download, 'punkt_tab', download_dir=nltk_data_dir, quiet=True)
     except Exception as e:
-        logger.error(f"Error downloading NLTK data: {str(e)}")
+        logger.error(f"Error handling NLTK data: {str(e)}")
         raise Exception(f"Failed to initialize application: {str(e)}")
 
     # Initialize the model and index

spaces_app.py

@@ -22,7 +22,9 @@ logger = logging.getLogger(__name__)
 
 # Initialize the FastAPI app in a separate thread
 def start_fastapi():
-    uvicorn.run(fastapi_app, host="0.0.0.0", port=8080)
+    # Don't run startup event here, as we'll run it manually
+    # This prevents duplicate initialization
+    uvicorn.run(fastapi_app, host="0.0.0.0", port=8080, lifespan="off")
 
 # Start FastAPI in a separate thread
 thread = Thread(target=start_fastapi, daemon=True)