fix method name

Author: wangzaijun

lightllm/server/router/model_infer/mode_backend/base_backend.py

@@ -559,7 +559,9 @@ def _get_classed_reqs(
         self._pre_handle_finished_reqs(finished_reqs=finished_reqs)
         # 如果使能了 cpu cache 功能，对于已经完成的请求，进行 gpu kv 卸载到 cpu cache的操作。
         if self.args.enable_cpu_cache:
-            true_finished_reqs = self.multi_level_cache_module.handle_finished_reqs(finished_reqs=finished_reqs)
+            true_finished_reqs = self.multi_level_cache_module.offload_finished_reqs_to_cpu_cache(
+                finished_reqs=finished_reqs
+            )
         else:
             true_finished_reqs = finished_reqs
 

lightllm/server/router/model_infer/mode_backend/multi_level_kv_cache.py

@@ -92,9 +92,9 @@ def load_cpu_cache_to_reqs(self, reqs: List[InferReq]):
             self.cpu_cache_client.lock.release()
         return
 
-    def handle_finished_reqs(self, finished_reqs: List[InferReq]) -> List[InferReq]:
+    def offload_finished_reqs_to_cpu_cache(self, finished_reqs: List[InferReq]) -> List[InferReq]:
         """
-        将满足cpu kv cache 卸载条件的请求进行处理，并返回需要真正退出的请求列表。
+        将满足cpu kv cache 卸载条件的请求进行处理, 并返回真的满足退出条件的请求list。
         """
 
         if self.args.enable_cpu_cache: