Update client.py

client.py

@@ -17,7 +17,7 @@ class Inference:
                 response.raise_for_status()
                 return response.content
             except requests.RequestException:
-                if response.status_code in {429, 500, 502, 503, 504}:
+                if e.response.status_code in {429, 500, 502, 503, 504}:
                     delay = min(2 ** retry, 60)
                     time.sleep(delay)
                 else: