updated app.py generate response function

app.py

@@ -12,6 +12,9 @@ model = AutoModelWithLMHead.from_pretrained("raghavdw/finedtuned_gpt2_medQA_mode
 
 def generate_query_response(prompt, max_length=200):
 
+    model = model
+    tokenizer = tokenizer
+
     input_ids = tokenizer.encode(prompt, return_tensors="pt")
 
     device = torch.device("cuda" if torch.cuda.is_available() else "cpu")