dssm train code

recommend-model-produce/src/main/python/models/dssm/net.py

@@ -100,18 +100,18 @@ class DSSMLayer(nn.Layer):
 
     def forward(self, left_features, right_features):
         # 获取两个视频的特征表示
-        paddle.static.Print(left_features, message="lqc left model input shape:")
-        paddle.static.Print(right_features, message="lqc right model input shape:")        
+        #paddle.static.Print(left_features, message="lqc left model input shape:")
+        #paddle.static.Print(right_features, message="lqc right model input shape:")        
         left_vec, right_vec = self.get_vectors(left_features, right_features)
-        paddle.static.Print(left_vec, message="lqc left model output shape:")
-        paddle.static.Print(right_vec, message="lqc right model output shape:")
+        #paddle.static.Print(left_vec, message="lqc left model output shape:")
+        #paddle.static.Print(right_vec, message="lqc right model output shape:")
         # 计算相似度
         sim_score = F.cosine_similarity(
             left_vec, 
             right_vec, 
             axis=1
         ).reshape([-1, 1])
-        paddle.static.Print(sim_score, message="lqc sim_score shape:")
+        #paddle.static.Print(sim_score, message="lqc sim_score shape:")
         return sim_score, left_vec, right_vec
 
     def get_vectors(self, left_features, right_features):

recommend-model-produce/src/main/python/models/dssm/static_model.py

@@ -57,9 +57,9 @@ class StaticModel():
             left_features, right_features = input
         else:
             label,left_features, right_features = input
-            paddle.static.Print(left_features, message="lqc left data feature shape:")
-            paddle.static.Print(right_features, message="lqc right data feature shape:")
-            paddle.static.Print(label, message="lqc label feature shape:")
+            #paddle.static.Print(left_features, message="lqc left data feature shape:")
+            #paddle.static.Print(right_features, message="lqc right data feature shape:")
+            #paddle.static.Print(label, message="lqc label feature shape:")
 
         # 获取相似度和特征向量
         sim_score, left_vec, right_vec = dssm_model(left_features, right_features)