Update app.py

app.py

@@ -9,9 +9,10 @@ UL2_API_URL = "https://api-inference.huggingface.co/models/google/flan-ul2"
 FLAN_API_URL = "https://api-inference.huggingface.co/models/google/flan-t5-xxl"
 
 headers = {"Authorization": f"Bearer {TOKEN}"}
+MAX_NEW_TOKENS = 128
 
 def query(text, api_url):
-    response = requests.post(api_url, headers=headers, json={"inputs":text})
+    response = requests.post(api_url, headers=headers, json={"inputs":text, "max_new_tokens":MAX_NEW_TOKENS})
     return response.json()