diff --git a/src/llm.cpp b/src/llm.cpp
index 8d32b45e..177c11ff 100644
--- a/src/llm.cpp
+++ b/src/llm.cpp
@@ -723,8 +723,8 @@ bool Llama2_7b::is_stop(int token_id) {
 
 std::vector<int> Qwen2_4b::tokenizer(const std::string& query) {
     auto ids = tokenizer_encode(query);
-    // auto prompt = "\n<|im_start|>user\n" + query + "<|im_end|>\n<|im_start|>assistant\n";
-    ids.insert(ids.begin(), {198, 151644, 872, 198});
+    // auto prompt = "<|im_start|>user\n" + query + "<|im_end|>\n<|im_start|>assistant\n";
+    ids.insert(ids.begin(), {151644, 872, 198});
     ids.insert(ids.end(), {151645, 198, 151644, 77091, 198});
     return ids;
 }
@@ -1321,4 +1321,4 @@ std::string Pipeline::build_prompt(const std::string& str) {
     prompt += assistant_ + "： ";
     return prompt;
 }
-// Pipeline end
\ No newline at end of file
+// Pipeline end