remove pre-cache

app.py

@@ -1,4 +1,6 @@
 import sys
+import os
+import shutil
 import warnings
 import spaces
 from threading import Thread
@@ -15,10 +17,6 @@ import math
 print(f"NumPy version: {np.__version__}")
 print(f"PyTorch version: {torch.__version__}")
 
-# cp the .gradio examples to /home/user/app/.gradio/
-import os
-import shutil
-
 # Suppress CUDA initialization warning
 warnings.filterwarnings("ignore", category=UserWarning, message="Can't initialize NVML")
 
@@ -191,4 +189,7 @@ if __name__ == "__main__":
         repo_id="orionweller/rank1-7b-awq",
     )
     print(f"Downloaded model to: {MODEL_PATH}")
+    # Create the cached_examples directory if it doesn't exist
+    os.makedirs("/home/user/app/.gradio/", exist_ok=True)
+    shutil.copytree("/home/user/app/.gradio/examples", "/home/user/app/.gradio/examples")
     demo.launch(share=True)