diff --git a/src/gausskernel/dbmind/tools/components/predictor/python/model.py b/src/gausskernel/dbmind/tools/components/predictor/python/model.py
index a3f6a741b..ab9619145 100644
--- a/src/gausskernel/dbmind/tools/components/predictor/python/model.py
+++ b/src/gausskernel/dbmind/tools/components/predictor/python/model.py
@@ -539,13 +539,14 @@ class RnnModel():
         keras.backend.clear_session()
         set_session(self.session)
         with self.graph.as_default():
+            # Judge whether the model needs to be initialized according to the changes of the model input and output dimensions.
             feature, label, need_init = self.parse(filename)
             os.environ['CUDA_VISIBLE_DEVICES'] = '0'
             epsilon = self.model_info.make_epsilon()
-            if need_init:
+            if need_init:# Cold start training
                 epoch_start = 0
                 self.model = self._build_model(epsilon)
-            else:
+            else:# Incremental training
                 epoch_start = int(self.model_info.last_epoch)
                 ratio_error = ratio_error_loss_wrapper(epsilon)
                 ratio_acc_2 = ratio_error_acc_wrapper(epsilon, 2)
@@ -556,12 +557,16 @@ class RnnModel():
             log_path = os.path.realpath(os.path.join(settings.PATH_LOG, self.model_info.model_name + '_log.json'))
             if not os.path.exists(log_path):
                 os.mknod(log_path, mode=0o600)
+            # Training logging callback function
             json_logging_callback = LossHistory(log_path, self.model_info.model_name, self.model_info.last_epoch)
+            # Data segmentation
             X_train, X_val, y_train, y_val = \
                 train_test_split(feature, label, test_size=0.1)
+            # model training
             self.model.fit(X_train, y_train, epochs=self.model_info.last_epoch,
                            batch_size=int(self.model_info.batch_size), validation_data=(X_val, y_val),
                            verbose=0, initial_epoch=epoch_start, callbacks=[json_logging_callback])
+            # save model
             self.model.save(self.model_info.model_path)
             val_pred = self.model.predict(X_val)
             val_re = get_ratio_errors_general(val_pred, y_val, epsilon)
diff --git a/src/gausskernel/dbmind/tools/components/sqldiag/algorithm/duration_time_model/template.py b/src/gausskernel/dbmind/tools/components/sqldiag/algorithm/duration_time_model/template.py
index ec20d0ede..41b000ace 100644
--- a/src/gausskernel/dbmind/tools/components/sqldiag/algorithm/duration_time_model/template.py
+++ b/src/gausskernel/dbmind/tools/components/sqldiag/algorithm/duration_time_model/template.py
@@ -27,6 +27,7 @@ from . import AbstractModel
 
 
 class TemplateModel(AbstractModel):
+    # Initialize algorithm parameters
     def __init__(self, params):
         super().__init__(params)
         self.bias = 1e-5