Update app.py

app.py

@@ -74,7 +74,7 @@ model = AutoModelForCausalLM.from_pretrained(
     torch_dtype=torch.bfloat16,
     device_map="auto"
 )
-end_of_sentence = tokenizer.convert_tokens_to_ids("<|im_end|>")
+end_of_sentence = tokenizer.convert_tokens_to_ids("<|eot_id|>")
 
 @spaces.GPU()
 def stream_chat(