Update app.py

app.py

@@ -111,7 +111,7 @@ def engine(input_embeds):
             batch_input_embeds = input_embeds[start:start+batch_size]
             outputs = m.generate(
                 inputs_embeds = batch_input_embeds.to("cuda"),
-                max_new_tokens = 512,
+                max_new_tokens = 16,
                 do_sample = True,
                 temperature = 0.6,
                 top_p = 0.9,
@@ -209,8 +209,8 @@ def chat(message, history, with_defense,threshold):
     messages = tok.apply_chat_template(chat, tokenize=False, add_generation_prompt=True)
     # Tokenize the messages string
     input_ids = tok([messages], return_tensors="pt")["input_ids"]
-    response= "[Gradient Cuff Checking: "+reject_information + "]\n"+ chat_engine(input_ids)
-    #response=chat_engine(input_ids)
+    #response= "[Gradient Cuff Checking: "+reject_information + "]\n"+ chat_engine(input_ids)
+    response=chat_engine(input_ids)
     response=response.split(" ")
 
     # Initialize an empty string to store the generated text