prepare train 40 epoch with this params.json

src/params.json

@@ -1,10 +1,10 @@
 {
   "bidirectional": true,
-  "use_residual": false,
-  "use_dropout": false,
-  "time_major": false,
+  "use_residual": true,
+  "use_dropout": true,
+  "time_major": true,
   "cell_type": "lstm",
-  "depth": 2,
+  "depth": 4,
   "attention_type": "Bahdanau",
   "hidden_units": 128,
   "optimizer": "adam",

src/test.py

@@ -29,8 +29,8 @@ def test(params):
     )
 
     # 读取模型路径
-    # save_path = './model/s2ss_chatbot_anti.ckpt'
-    save_path = './model/s2ss_chatbot.ckpt'
+    save_path = './model/s2ss_chatbot_anti.ckpt'
+    # save_path = './model/s2ss_chatbot.ckpt'
 
     tf.reset_default_graph()
     model_pred = SequenceToSequence(

src/train_anti.py

@@ -20,7 +20,7 @@ def test(params):
     x_data, y_data = pickle.load(open('chatbot.pkl', 'rb'))
     ws = pickle.load(open('ws.pkl', 'rb'))
 
-    n_epoch = 2  # 训练轮次
+    n_epoch = 40  # 训练轮次
     batch_size = 128
     steps = int(len(x_data) / batch_size) + 1
 

src/web.py

@@ -1,6 +1,6 @@
 # -*- coding: utf-8 -*-
 """
-   Description :   train_anti 测试
+   Description :   web interface
    Author :        xxm
 """
 import sys
@@ -18,9 +18,6 @@ def test(params, infos):
     x_data, _ = pickle.load(open('chatbot.pkl', 'rb'))
     ws = pickle.load(open('ws.pkl', 'rb'))
 
-    for x in x_data[:5]:
-        print(' '.join(x))
-
     config = tf.ConfigProto(
         device_count={'CPU': 1, 'GPU': 0},
         allow_soft_placement=True,
@@ -50,17 +47,13 @@ def test(params, infos):
             x, xl = next(bar)
             x = np.flip(x, axis=1)
 
-            print(x, xl)
-
             pred = model_pred.predict(
                 sess,
                 np.array(x),
                 np.array(xl)
             )
             print(pred)
 
-            print(ws.inverse_transform(x[0]))
-
             for p in pred:
                 ans = ws.inverse_transform(p)
                 print(ans)