diff --git a/classifer.py b/classifer.py
index fc2a5a1..0805821 100644
--- a/classifer.py
+++ b/classifer.py
@@ -97,7 +97,7 @@ class WoodClass(object):
             sample = cv2.resize(x, (self.ww, self.hh))
         return sample
 
-    def fit_pictures(self, data_path=ROOT_DIR):
+    def fit_pictures(self, data_path=ROOT_DIR, file_name=None):
         """
         根据给出的data_path 进行 fit.如果没有给出data目录，那么将会使用当前文件夹
         :param data_path:
@@ -110,7 +110,7 @@ class WoodClass(object):
         x, y = result
         score = self.fit(x, y)
         print('model score', score)
-        model_name = self.save()
+        model_name = self.save(file_name)
         return model_name
 
     def fit(self, x,  y, test_size=0.1):
diff --git a/socket_detector.py b/socket_detector.py
index a9468b1..4ce2e05 100644
--- a/socket_detector.py
+++ b/socket_detector.py
@@ -33,8 +33,12 @@ def process_cmd(cmd: str, data: any, connected_sock: socket.socket, detector: Wo
         response = simple_sock(connected_sock, cmd_type=cmd, result=wood_color)
     elif cmd == 'TR':
         detector = WoodClass(w=4096, h=1200, n=3000, debug_mode=False)
+        model_name = None
+        if "/n" in data:
+            data, model_name = data.split("/n", 1)
+            model_name = model_name + ".p"
         settings.data_path = data
-        settings.model_path = ROOT_DIR / 'models' / detector.fit_pictures(data_path=settings.data_path)
+        settings.model_path = ROOT_DIR / 'models' / detector.fit_pictures(data_path=settings.data_path, file_name=model_name)
         response = simple_sock(connected_sock, cmd_type=cmd)
     elif cmd == 'MD':
         settings.model_path = data