Update app.py

app.py

@@ -51,7 +51,7 @@ def generate_responses(prompt):
 
     # Common generation parameters
     generation_kwargs = {
-        "max_new_tokens": 100,
+        "max_new_tokens": 120,
         "num_return_sequences": 1,
         "pad_token_id": sft_tokenizer.eos_token_id, # Can use either tokenizer's pad token
         "top_k": 50,