Update app.py

app.py

@@ -138,7 +138,7 @@ def reply(message: str, history: list[str]) -> str:
     text = tokenizer.apply_chat_template(
         history_transformer_format, tokenize=False, add_generation_prompt=True
     )
-    model_inputs = tokenizer([text], return_tensors="pt").to("cuda:0")
+    model_inputs = tokenizer([text], return_tensors="pt").to("cuda")
 
     generate_kwargs = dict(model_inputs, streamer=streamer, max_new_tokens=512)
     t = threading.Thread(target=chatmodel.generate, kwargs=generate_kwargs)