Update app.py

app.py

@@ -4,7 +4,11 @@ import os
 
 # Load token and create client
 token = os.getenv("huggingface_token")
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta", token=token)
+client = InferenceClient(
+    "HuggingFaceH4/zephyr-7b-beta",
+    token=token,
+    provider="auto"  # <- fixes the issue
+)
 
 # Load system messages from environment
 SYSTEM_MESSAGES = {