diff --git a/.idea/workspace.xml b/.idea/workspace.xml
index db98590..18436f8 100644
--- a/.idea/workspace.xml
+++ b/.idea/workspace.xml
@@ -181,7 +181,7 @@
   <component name="PropertiesComponent">
     <property name="WebServerToolWindowFactoryState" value="false" />
     <property name="aspect.path.notification.shown" value="true" />
-    <property name="com.android.tools.idea.instantapp.provision.ProvisionBeforeRunTaskProvider.myTimeStamp" value="1588169908975" />
+    <property name="com.android.tools.idea.instantapp.provision.ProvisionBeforeRunTaskProvider.myTimeStamp" value="1588169942720" />
     <property name="go.gopath.indexing.explicitly.defined" value="true" />
     <property name="nodejs_interpreter_path.stuck_in_default_project" value="undefined stuck path" />
     <property name="nodejs_npm_path_reset_for_default_project" value="true" />
@@ -190,6 +190,12 @@
     <property name="project.structure.side.proportion" value="0.0" />
     <property name="settings.editor.selected.configurable" value="http.proxy" />
   </component>
+  <component name="RecentsManager">
+    <key name="MoveFile.RECENT_KEYS">
+      <recent name="$PROJECT_DIR$" />
+      <recent name="$PROJECT_DIR$/model" />
+    </key>
+  </component>
   <component name="RunDashboard">
     <option name="ruleStates">
       <list>
diff --git a/main.py b/main.py
index b1e50ed..283791d 100644
--- a/main.py
+++ b/main.py
@@ -4,9 +4,11 @@ import os
 from sklearn.model_selection import train_test_split
 from model_tensorflow import train, predict
 
+frame = "tensorflow"
+
 
 class Config:
-	feature_columns = list([2,5])
+	feature_columns = list([2, 5])
 	label_columns = [5]
 	feature_and_label_columns = feature_columns + label_columns
 	label_in_feature_columns = (lambda x, y: [x.index(i) for i in y])(feature_columns, label_columns)
@@ -51,9 +53,9 @@ class Config:
 	if not os.path.exists(figure_save_path):
 		os.mkdir(figure_save_path)
 
-	#used_frame = frame
-	#model_postfix = {"pytorch": ".pth", "keras": ".h5", "tensorflow": ".ckpt"}
-	#model_name = "model_" + continue_flag + used_frame + model_postfix[used_frame]
+	used_frame = frame
+	model_postfix = {"pytorch": ".pth", "keras": ".h5", "tensorflow": ".ckpt"}
+	model_name = "model_" + continue_flag + used_frame + model_postfix[used_frame]
 
 
 class Data:
diff --git a/serve.py b/serve.py
index 1bce038..c70b597 100644
--- a/serve.py
+++ b/serve.py
@@ -5,14 +5,159 @@ from http.server import BaseHTTPRequestHandler, HTTPServer
 import cgi
 import json
 from urllib import parse
+import pandas as pd
+import numpy as np
+import os
+from sklearn.model_selection import train_test_split
+from model_tensorflow import train, predict
+
+
+class Config:
+	feature_columns = list([2, 5])
+	label_columns = [5]
+	feature_and_label_columns = feature_columns + label_columns
+	label_in_feature_columns = (lambda x, y: [x.index(i) for i in y])(feature_columns, label_columns)
+
+	predict_day = 1
+
+	input_size = len(feature_columns)
+	output_size = len(label_columns)
+
+	hidden_size = 128
+	lstm_layers = 2
+	dropout_rate = 0.2
+	time_step = 20
+
+	do_train = True
+	do_predict = True
+	add_train = False
+	shuffle_train_data = True
+
+	train_data_rate = 0.95
+	valid_data_rate = 0.15
+
+	batch_size = 64
+	learning_rate = 0.001
+	epoch = 20
+	patience = 5
+	random_seed = 42
+
+	do_continue_train = False
+	continue_flag = ""
+	if do_continue_train:
+		shuffle_train_data = False
+		batch_size = 1
+		continue_flag = "continue_"
+
+	train_data_path = "./data/stock_data.csv"
+	model_save_path = "./checkpoint/"
+	figure_save_path = "./figure/"
+	do_figure_save = False
+	if not os.path.exists(model_save_path):
+		os.mkdir(model_save_path)
+	if not os.path.exists(figure_save_path):
+		os.mkdir(figure_save_path)
+
+	used_frame = frame
+	model_postfix = {"pytorch": ".pth", "keras": ".h5", "tensorflow": ".ckpt"}
+	model_name = "model_" + continue_flag + used_frame + model_postfix[used_frame]
+
+
+class Data:
+	def __init__(self, config):
+		self.config = config
+		self.data, self.data_column_name = self.read_data()
+
+		self.data_num = self.data.shape[0]
+		self.train_num = int(self.data_num * self.config.train_data_rate)
+
+		self.mean = np.mean(self.data, axis=0)
+		self.std = np.std(self.data, axis=0)
+		self.norm_data = (self.data - self.mean) / self.std
+
+		self.start_num_in_test = 0
+
+	def read_data(self):
+		init_data = pd.read_csv(self.config.train_data_path,
+		                        usecols=self.config.feature_and_label_columns)
+		return init_data.values, init_data.columns.tolist()
+
+	def get_train_and_valid_data(self):
+		feature_data = self.norm_data[:self.train_num]
+		label_data = self.norm_data[self.config.predict_day: self.config.predict_day + self.train_num,
+		             self.config.label_in_feature_columns]
+		if not self.config.do_continue_train:
+			train_x = [feature_data[i:i + self.config.time_step] for i in range(self.train_num - self.config.time_step)]
+			train_y = [label_data[i:i + self.config.time_step] for i in range(self.train_num - self.config.time_step)]
+		else:
+			train_x = [
+				feature_data[start_index + i * self.config.time_step: start_index + (i + 1) * self.config.time_step]
+				for start_index in range(self.config.time_step)
+				for i in range((self.train_num - start_index) // self.config.time_step)]
+			train_y = [
+				label_data[start_index + i * self.config.time_step: start_index + (i + 1) * self.config.time_step]
+				for start_index in range(self.config.time_step)
+				for i in range((self.train_num - start_index) // self.config.time_step)]
+
+		train_x, train_y = np.array(train_x), np.array(train_y)
+
+		train_x, valid_x, train_y, valid_y = train_test_split(train_x, train_y, test_size=self.config.valid_data_rate,
+		                                                      random_state=self.config.random_seed,
+		                                                      shuffle=self.config.shuffle_train_data)
+		return train_x, valid_x, train_y, valid_y
+
+	def get_test_data(self, return_label_data=False):
+		feature_data = self.norm_data[self.train_num:]
+		self.start_num_in_test = feature_data.shape[0] % self.config.time_step
+		time_step_size = feature_data.shape[0] // self.config.time_step
+
+		test_x = [feature_data[self.start_num_in_test + i * self.config.time_step: self.start_num_in_test + (
+				i + 1) * self.config.time_step]
+		          for i in range(time_step_size)]
+		if return_label_data:
+			label_data = self.norm_data[self.train_num + self.start_num_in_test:, self.config.label_in_feature_columns]
+			return np.array(test_x), label_data
+		return np.array(test_x)
+
+
+def draw(config, origin_data, predict_norm_data):
+	label_norm_data = origin_data.norm_data[origin_data.train_num + origin_data.start_num_in_test:,
+	                  config.label_in_feature_columns]
+	assert label_norm_data.shape[0] == predict_norm_data.shape[
+		0], "The element number in origin and predicted data is different"
+
+	label_name = [origin_data.data_column_name[i] for i in config.label_in_feature_columns]
+	label_column_num = len(config.label_columns)
+
+	loss = np.mean((label_norm_data[config.predict_day:] - predict_norm_data[:-config.predict_day]) ** 2, axis=0)
+	print("The mean squared error of stock {} is ".format(label_name), loss)
+
+	label_X = range(origin_data.data_num - origin_data.train_num - origin_data.start_num_in_test)
+	predict_X = [x + config.predict_day for x in label_X]
+
+	label_data = label_norm_data * origin_data.std[config.label_in_feature_columns] + \
+	             origin_data.mean[config.label_in_feature_columns]
+
+	predict_data = predict_norm_data * origin_data.std[config.label_in_feature_columns] + \
+	               origin_data.mean[config.label_in_feature_columns]
+
+	print(label_data)
+	print(predict_data)
+
 
 PORT_NUMBER = 8000
-
 lock = Lock()
+config = Config()
+frame = "tensorflow"
 
 
 def train_models():
 	lock.acquire()
+	np.random.seed(config.random_seed)
+	data_gainer = Data(config)
+
+	train_X, valid_X, train_Y, valid_Y = data_gainer.get_train_and_valid_data()
+	train(config, train_X, train_Y, valid_X, valid_Y)
 
 	lock.release()
 
@@ -34,6 +179,25 @@ class MyHandler(BaseHTTPRequestHandler):
 				job = query.get('job')[0]
 				gpu_model = query.get('gpu_model')[0]
 				time = query.get('time')[0]
+				data_gainer = Data(config)
+				test_X, test_Y = data_gainer.get_test_data(return_label_data=True)
+				pred_result = predict(config, test_X)
+				draw(config, data_gainer, pred_result)
+				msg = {'code': 1, 'error': "container not exist"}
+			except Exception as e:
+				msg = {'code': 2, 'error': str(e)}
+			self.send_response(200)
+			self.send_header('Content-type', 'application/json')
+			self.end_headers()
+			self.wfile.write(bytes(json.dumps(msg), "utf-8"))
+
+		elif req.path == "/train":
+			try:
+				job = query.get('job')[0]
+				gpu_model = query.get('gpu_model')[0]
+				time = query.get('time')[0]
+				t = Thread(target=train_models(), name='train_models', args=(job,))
+				t.start()
 				msg = {'code': 1, 'error': "container not exist"}
 			except Exception as e:
 				msg = {'code': 2, 'error': str(e)}