From 345f788a4faf8e931866c6bfae2e541c90269641 Mon Sep 17 00:00:00 2001
From: chenjian <chenjian26@baidu.com>
Date: Thu, 14 Dec 2023 14:26:05 +0800
Subject: [PATCH] Add detailed log for llm service

---
 llm/fastdeploy_llm/serving/triton_model.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/llm/fastdeploy_llm/serving/triton_model.py b/llm/fastdeploy_llm/serving/triton_model.py
index f76d061192..bae3c5c619 100644
--- a/llm/fastdeploy_llm/serving/triton_model.py
+++ b/llm/fastdeploy_llm/serving/triton_model.py
@@ -45,8 +45,8 @@ def stream_call_back(call_back_task, token_tuple, index, is_last_token,
         "OUT", np.array(
             [json.dumps(out)], dtype=np.object_))
     if is_last_token:
-        all_token_ids = [t[0] for t in call_back_task.result.completion_tokens]
-        all_strs = "".join[t[1] for t in call_back_task.result.completion_tokens]   
+        all_token_ids = [t[0] for t in call_back_task.result.completion_tokens] + [token_tuple[0]]
+        all_strs = "".join[t[1] for t in call_back_task.result.completion_tokens] + [token_tuple[1]] 
         logger.info("Model output for req_id: {}  results_all: {} tokens_all: {}".format(call_back_task.task_id, all_strs, all_token_ids))
         sender[call_back_task.task_id].send(
             pb_utils.InferenceResponse([out_tensor]),