Update app.py

app.py

@@ -128,10 +128,19 @@ async def gen_show(script):
 def generator(messages):
     input_ids = tokenizer.apply_chat_template(
         conversation=messages,
+        add_generation_prompt=True,
         tokenize=True,
         return_tensors='pt'
     )
-    output_ids = model.generate(input_ids.to('cuda'), eos_token_id=tokenizer.eos_token_id)
+    
+    output_ids = model.generate(
+        input_ids.to('cuda'), 
+        eos_token_id=tokenizer.eos_token_id,
+        max_new_tokens=4096,
+        temperature=0.5,
+        repetition_penalty=1.2,
+        )
+    
     results = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=True)
     return results