Update app.py

app.py

@@ -20,7 +20,7 @@ if torch.cuda.is_available():
     model_id = "Qwen/Qwen2.5-7B-Instruct"
     #model_id = "BenBranyon/sumbot7b"
     model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")
-    tokenizer = AutoTokenizer.from_pretrained(model_id)
+    tokenizer = AutoTokenizer.from_pretrained(model_id, use_fast=True, model_max_length=MAX_INPUT_TOKEN_LENGTH, padding="longest", language="en")
     tokenizer.use_default_system_prompt = False
 
 #Inference API Code