diff --git a/examples/benchmarks/GATs/workflow_config_gats.yaml b/examples/benchmarks/GATs/workflow_config_gats.yaml
index c38b4b3129..ef2bbd1864 100644
--- a/examples/benchmarks/GATs/workflow_config_gats.yaml
+++ b/examples/benchmarks/GATs/workflow_config_gats.yaml
@@ -53,6 +53,8 @@ task:
             metric: loss
             loss: mse
             base_model: LSTM
+            with_pretrain: True
+            model_path: "benchmarks/LSTM/model_lstm_csi300.pkl"
             seed: 0
             GPU: 0
     dataset:
diff --git a/qlib/contrib/data/handler.py b/qlib/contrib/data/handler.py
index e97b00c241..6b3de336dc 100644
--- a/qlib/contrib/data/handler.py
+++ b/qlib/contrib/data/handler.py
@@ -32,79 +32,6 @@ def check_transform_proc(proc_l, fit_start_time, fit_end_time):
     return new_l
 
 
-class ALPHA360_Denoise(DataHandlerLP):
-    def __init__(self, instruments="csi500", start_time=None, end_time=None, fit_start_time=None, fit_end_time=None):
-        data_loader = {
-            "class": "QlibDataLoader",
-            "kwargs": {
-                "config": {
-                    "feature": self.get_feature_config(),
-                    "label": self.get_label_config(),
-                },
-            },
-        }
-
-        learn_processors = [
-            {"class": "DropnaLabel", "kwargs": {"fields_group": "label"}},
-            {"class": "CSZScoreNorm", "kwargs": {"fields_group": "label"}},
-        ]
-        infer_processors = [
-            {"class": "ProcessInf", "kwargs": {}},
-            {"class": "TanhProcess", "kwargs": {}},
-            {"class": "Fillna", "kwargs": {}},
-        ]
-
-        super().__init__(
-            instruments,
-            start_time,
-            end_time,
-            data_loader=data_loader,
-            learn_processors=learn_processors,
-            infer_processors=infer_processors,
-        )
-
-    def get_label_config(self):
-        return (["Ref($close, -2)/Ref($close, -1) - 1"], ["LABEL0"])
-
-    def get_feature_config(self):
-
-        fields = []
-        names = []
-
-        for i in range(59, 0, -1):
-            fields += ["Ref($close, %d)/$close" % (i)]
-            names += ["CLOSE%d" % (i)]
-        fields += ["$close/$close"]
-        names += ["CLOSE0"]
-        for i in range(59, 0, -1):
-            fields += ["Ref($open, %d)/$close" % (i)]
-            names += ["OPEN%d" % (i)]
-        fields += ["$open/$close"]
-        names += ["OPEN0"]
-        for i in range(59, 0, -1):
-            fields += ["Ref($high, %d)/$close" % (i)]
-            names += ["HIGH%d" % (i)]
-        fields += ["$high/$close"]
-        names += ["HIGH0"]
-        for i in range(59, 0, -1):
-            fields += ["Ref($low, %d)/$close" % (i)]
-            names += ["LOW%d" % (i)]
-        fields += ["$low/$close"]
-        names += ["LOW0"]
-        for i in range(59, 0, -1):
-            fields += ["Ref($vwap, %d)/$close" % (i)]
-            names += ["VWAP%d" % (i)]
-        fields += ["$vwap/$close"]
-        names += ["VWAP0"]
-        for i in range(59, 0, -1):
-            fields += ["Ref($volume, %d)/$volume" % (i)]
-            names += ["VOLUME%d" % (i)]
-        fields += ["$volume/$volume"]
-        names += ["VOLUME0"]
-
-        return fields, names
-
-
 _DEFAULT_LEARN_PROCESSORS = [
     {"class": "DropnaLabel"},
     {"class": "CSZScoreNorm", "kwargs": {"fields_group": "label"}},
diff --git a/qlib/contrib/model/pytorch_alstm.py b/qlib/contrib/model/pytorch_alstm.py
index 40c2f82266..ba75c70caa 100644
--- a/qlib/contrib/model/pytorch_alstm.py
+++ b/qlib/contrib/model/pytorch_alstm.py
@@ -113,6 +113,9 @@ def __init__(
             )
         )
 
+        np.random.seed(self.seed)
+        torch.manual_seed(self.seed)
+
         self.ALSTM_model = ALSTMModel(
             d_feat=self.d_feat,
             hidden_size=self.hidden_size,
diff --git a/qlib/contrib/model/pytorch_gats.py b/qlib/contrib/model/pytorch_gats.py
index e9cbcf9cb0..2e19c0542c 100644
--- a/qlib/contrib/model/pytorch_gats.py
+++ b/qlib/contrib/model/pytorch_gats.py
@@ -9,10 +9,15 @@
 import numpy as np
 import pandas as pd
 import copy
-from ...utils import create_save_path
-from ...log import get_module_logger
-
-
+from sklearn.metrics import roc_auc_score, mean_squared_error
+import logging
+from ...utils import (
+    unpack_archive_with_buffer,
+    save_multiple_parts_file,
+    create_save_path,
+    drop_nan_by_y_index,
+)
+from ...log import get_module_logger, TimeInspector
 import torch
 import torch.nn as nn
 import torch.optim as optim
@@ -54,6 +59,7 @@ def __init__(
         loss="mse",
         base_model="GRU",
         with_pretrain=True,
+        model_path=None,
         optimizer="adam",
         GPU="0",
         seed=0,
@@ -76,6 +82,7 @@ def __init__(
         self.loss = loss
         self.base_model = base_model
         self.with_pretrain = with_pretrain
+        self.model_path = model_path
         self.visible_GPU = GPU
         self.use_gpu = torch.cuda.is_available()
         self.seed = seed
@@ -94,6 +101,7 @@ def __init__(
             "\nloss_type : {}"
             "\nbase_model : {}"
             "\nwith_pretrain : {}"
+            "\nmodel_path : {}"
             "\nvisible_GPU : {}"
             "\nuse_GPU : {}"
             "\nseed : {}".format(
@@ -109,12 +117,14 @@ def __init__(
                 loss,
                 base_model,
                 with_pretrain,
+                model_path,
                 GPU,
                 self.use_gpu,
                 seed,
             )
         )
-
+        np.random.seed(self.seed)
+        torch.manual_seed(self.seed)
         self.GAT_model = GATModel(
             d_feat=self.d_feat,
             hidden_size=self.hidden_size,
@@ -254,14 +264,17 @@ def fit(
 
         # load pretrained base_model
         if self.with_pretrain:
+            if self.model_path == None:
+                raise ValueError("the path of the pretrained model should be given first!")
             self.logger.info("Loading pretrained model...")
             if self.base_model == "LSTM":
                 pretrained_model = LSTMModel()
-                pretrained_model.load_state_dict(torch.load("benchmarks/LSTM/model_lstm_csi300.pkl"))
-
+                pretrained_model.load_state_dict(torch.load(self.model_path))
             elif self.base_model == "GRU":
                 pretrained_model = GRUModel()
-                pretrained_model.load_state_dict(torch.load("benchmarks/GRU/model_gru_csi300.pkl"))
+                pretrained_model.load_state_dict(torch.load(self.model_path))
+            else:
+                raise ValueError("unknown base model name `%s`" % self.base_model)
 
             model_dict = self.GAT_model.state_dict()
             pretrained_dict = {k: v for k, v in pretrained_model.state_dict().items() if k in model_dict}
diff --git a/qlib/contrib/model/pytorch_gru.py b/qlib/contrib/model/pytorch_gru.py
index 5daf4707ed..4ef4e25531 100755
--- a/qlib/contrib/model/pytorch_gru.py
+++ b/qlib/contrib/model/pytorch_gru.py
@@ -113,6 +113,9 @@ def __init__(
             )
         )
 
+        np.random.seed(self.seed)
+        torch.manual_seed(self.seed)
+
         self.gru_model = GRUModel(
             d_feat=self.d_feat,
             hidden_size=self.hidden_size,
diff --git a/qlib/contrib/model/pytorch_lstm.py b/qlib/contrib/model/pytorch_lstm.py
index eef1680ecd..59df51e427 100755
--- a/qlib/contrib/model/pytorch_lstm.py
+++ b/qlib/contrib/model/pytorch_lstm.py
@@ -113,6 +113,9 @@ def __init__(
             )
         )
 
+        np.random.seed(self.seed)
+        torch.manual_seed(self.seed)
+
         self.lstm_model = LSTMModel(
             d_feat=self.d_feat,
             hidden_size=self.hidden_size,
diff --git a/qlib/contrib/model/pytorch_nn.py b/qlib/contrib/model/pytorch_nn.py
index d324e27aac..a60227ff0e 100644
--- a/qlib/contrib/model/pytorch_nn.py
+++ b/qlib/contrib/model/pytorch_nn.py
@@ -61,6 +61,7 @@ def __init__(
         optimizer="gd",
         loss="mse",
         GPU="0",
+        seed=0,
         **kwargs
     ):
         # Set logger.
@@ -80,6 +81,7 @@ def __init__(
         self.loss_type = loss
         self.visible_GPU = GPU
         self.use_GPU = torch.cuda.is_available()
+        self.seed = seed
 
         self.logger.info(
             "DNN parameters setting:"
@@ -94,6 +96,7 @@ def __init__(
             "\noptimizer : {}"
             "\nloss_type : {}"
             "\neval_steps : {}"
+            "\nseed : {}"
             "\nvisible_GPU : {}"
             "\nuse_GPU : {}".format(
                 layers,
@@ -107,11 +110,15 @@ def __init__(
                 optimizer,
                 loss,
                 eval_steps,
+                seed,
                 GPU,
                 self.use_GPU,
             )
         )
 
+        np.random.seed(self.seed)
+        torch.manual_seed(self.seed)
+
         if loss not in {"mse", "binary"}:
             raise NotImplementedError("loss {} is not supported!".format(loss))
         self._scorer = mean_squared_error if loss == "mse" else roc_auc_score
diff --git a/qlib/contrib/model/pytorch_sfm.py b/qlib/contrib/model/pytorch_sfm.py
index 228c0aee5d..d845f62452 100644
--- a/qlib/contrib/model/pytorch_sfm.py
+++ b/qlib/contrib/model/pytorch_sfm.py
@@ -282,6 +282,9 @@ def __init__(
             )
         )
 
+        np.random.seed(self.seed)
+        torch.manual_seed(self.seed)
+
         self.sfm_model = SFM_Model(
             d_feat=self.d_feat,
             output_dim=self.output_dim,