-
Notifications
You must be signed in to change notification settings - Fork 5
/
Copy pathbert_method.py
195 lines (168 loc) · 8.77 KB
/
bert_method.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Mon Oct 14 23:36:17 2019
@author: chenjiannan
"""
import numpy as np
import pandas as pd
from keras_bert import load_trained_model_from_checkpoint
from sklearn.model_selection import train_test_split
import pickle
from keras.layers import Input,Dense,SpatialDropout1D,Bidirectional,LSTM,\
GlobalAveragePooling1D, Concatenate, GlobalMaxPooling1D,\
Lambda, CuDNNLSTM,CuDNNGRU
from keras.callbacks import EarlyStopping, ModelCheckpoint
from keras.models import Model
from keras import backend as K
from keras import optimizers
from my_utils import f1
def model_bert_txt(config_path,checkpoint_path,metric = f1, max_txt_len = 100):
bert_model = load_trained_model_from_checkpoint(config_path,
checkpoint_path,
seq_len=None)
inp_txt_x1 = Input(shape=(max_txt_len,))
inp_txt_x2 = Input(shape=(max_txt_len,))
for i in range(20):
bert_model.layers[-i].trainable
x1 = bert_model([inp_txt_x1, inp_txt_x2])
x1 = Lambda(lambda x: x)(x1)
x1 = SpatialDropout1D(0.3)(x1)
max_pool = GlobalMaxPooling1D()(x1)
avg_pool = GlobalAveragePooling1D()(x1)
pools = Concatenate()([max_pool, avg_pool])
predictions = Dense(1, activation='sigmoid')(pools)
model = Model(inputs=[inp_txt_x1, inp_txt_x2],
outputs=predictions)
adam = optimizers.Adam(lr=learning_rate)
model.compile(optimizer=adam, loss='binary_crossentropy', metrics=[metric])
return model
def model_bert_txt_lstm(config_path,checkpoint_path,metric = f1, max_txt_len = 100):
bert_model = load_trained_model_from_checkpoint(config_path,
checkpoint_path,
seq_len=None)
inp_txt_x1 = Input(shape=(max_txt_len,))
inp_txt_x2 = Input(shape=(max_txt_len,))
for i in range(20):
bert_model.layers[-i].trainable
x1 = bert_model([inp_txt_x1, inp_txt_x2])
x1 = Lambda(lambda x: x)(x1)
x1 = SpatialDropout1D(0.3)(x1)
max_pool = GlobalMaxPooling1D()(x1)
avg_pool = GlobalAveragePooling1D()(x1)
pools = Concatenate()([max_pool, avg_pool])
predictions = Dense(1, activation='sigmoid')(pools)
model = Model(inputs=[inp_txt_x1, inp_txt_x2],
outputs=predictions)
adam = optimizers.Adam(lr=learning_rate)
model.compile(optimizer=adam, loss='binary_crossentropy', metrics=[metric])
return model
def model_bert_v1(config_path,checkpoint_path,metric = f1):
bert_model = load_trained_model_from_checkpoint(config_path,
checkpoint_path,
seq_len=None)
inp_txt_x1 = Input(shape=(None,))
inp_txt_x2 = Input(shape=(None,))
inp_entity_x1 = Input(shape=(None,))
inp_entity_x2 = Input(shape=(None,))
input_entity_txt = Concatenate()([inp_txt_x1, inp_entity_x1])
input_entity_txt2 = Concatenate()([inp_txt_x2, inp_entity_x2])
for i in range(10):
bert_model.layers[-i].trainable
x_entity = bert_model([inp_entity_x1, inp_entity_x2])
x_entity = Lambda(lambda x:x)(x_entity)
x1 = bert_model([input_entity_txt, input_entity_txt2])
x1 = Lambda(lambda x: x)(x1)
x1 = SpatialDropout1D(0.3)(x1)
bilstm = Bidirectional(CuDNNLSTM(256, return_sequences=True, ))
x1 = bilstm(x1)
x_entity = bilstm(x_entity)
max_pool1 = GlobalMaxPooling1D()(x1)
avg_pool1 = GlobalAveragePooling1D()(x1)
max_pool2 = GlobalMaxPooling1D()(x_entity)
avg_pool2 = GlobalAveragePooling1D()(x_entity)
pools1 = Concatenate()([max_pool1, avg_pool1,])
pools2 = Concatenate()([max_pool2, avg_pool2,])
pools = Concatenate()([pools1, pools2])
predictions = Dense(1, activation='sigmoid')(pools)
model = Model(inputs=[inp_txt_x1, inp_txt_x2,
inp_entity_x1, inp_entity_x2],
outputs=predictions)
adam = optimizers.Adam(lr=learning_rate)
model.compile(optimizer=adam, loss='binary_crossentropy', metrics=[metric])
return model
if __name__ == '__main__':
config_path = './chinese_L-12_H-768_A-12/bert_config.json'
checkpoint_path = './chinese_L-12_H-768_A-12/bert_model.ckpt'
## 1.1 读取文件
with open('train_data_model_bert_char.pkl', 'rb') as f:
train_data = pickle.load(f)
random_state = 10
learning_rate = 0.01
## 1.2 load data 每次运行
y = train_data['y_train']
y_entity = train_data['y_entity_negative']
X_train_txt = train_data['X_train_txt']
X_train_title = train_data['X_train_title']
target_entity_train = train_data['target_entity_train']
## 1.3 记录title是否为空
isnotitle2 =(X_train_title.sum(axis=1)<1)
### 只使用title非空的部分
X_train_title_nonan = X_train_title[~isnotitle2, :]
y_title_nonan = y[~isnotitle2]
y_title_nonan_entity = y_entity[~isnotitle2]
y_index = np.array(range(0,y_title_nonan.shape[0]))
title_X_train_nonan, title_X_test_nonan, y_train_index, y_test_index = train_test_split(X_train_title_nonan, y_index,
test_size = 0.1,
random_state = random_state)
### 获取句子的负面消息 和entity的负面消息
title_y_train_nonan, title_y_test_nonan = y_title_nonan[y_train_index], y_title_nonan[y_test_index]
title_y_train_nonan_entity, title_y_test_nonan_entity = y_title_nonan_entity[y_train_index], y_title_nonan_entity[y_test_index]
X_train_txt_cut = X_train_txt[~isnotitle2,:]
target_entity_train_cut = target_entity_train[~isnotitle2]
txt_X_train_nonan, txt_X_test_nonan = X_train_txt_cut[y_train_index, :], X_train_txt_cut[y_test_index, :]
target_entity_train_cut_train_nonan, target_entity_train_cut_test_nonan = target_entity_train_cut[y_train_index],target_entity_train_cut[y_test_index]
## 1.4 have no title
### only txt
y_index = np.array(range(0,X_train_txt.shape[0]))
X_txt_train, X_txt_test, y_train_index, y_test_index = train_test_split(X_train_txt, y_index,
test_size = 0.1,
random_state = random_state)
y_train, y_test = y[y_train_index], y[y_test_index]
y_entity_train, y_entity_test = y_entity[y_train_index], y_entity[y_test_index]
X_target_entity_train, X_target_entity_test = target_entity_train[y_train_index],target_entity_train[y_test_index]
#%% 2.load test data, 载入测试集数据，用于预测
with open('test_data_model_bert_char.pkl', 'rb') as f:
test_data = pickle.load(f)
X_test_txt = test_data['X_test_txt']
X_test_title = test_data['X_test_title']
target_entity_test = test_data['target_entity_test']
#%% 3.建立bert模型，训练
config_path = './chinese_L-12_H-768_A-12/bert_config.json'
checkpoint_path = './chinese_L-12_H-768_A-12/bert_model.ckpt'
cut_shape = 500
model = model_bert_txt(config_path,checkpoint_path,metric = f1,
max_txt_len = X_txt_train[:,:cut_shape].shape[1])
filepath = "best_weights.h5"
early_stopping = EarlyStopping(monitor='val_loss',patience=2)
checkpoint = ModelCheckpoint(filepath, monitor='val_acc', verbose=0, save_best_only=True, mode='max', period=1)
callbacks_list = [checkpoint,early_stopping]
model.fit([X_txt_train[:,:cut_shape], np.zeros(X_txt_train[:,:cut_shape].shape)],
y_train,
batch_size = 16,
epochs = 20,
validation_data = [[X_txt_test[:,:cut_shape], np.zeros(X_txt_test[:,:cut_shape].shape)],
y_test],
callbacks=callbacks_list)
early_stopping = EarlyStopping(monitor='val_loss',patience=5)
model_bert_ = model_bert_v1(config_path, checkpoint_path,metric = f1)
cut_shape = 450
model_bert_.fit([X_txt_train[:,:cut_shape], np.zeros(X_txt_train[:,:cut_shape].shape),
X_target_entity_train[:,:cut_shape], np.zeros(X_target_entity_train[:,:cut_shape].shape)],
y_entity_train,
batch_size = 32,
epochs = 20,
validation_data = [[X_txt_test[:,:cut_shape], np.zeros(X_txt_test[:,:cut_shape].shape),
X_target_entity_test[:,:cut_shape], np.zeros(X_target_entity_test[:,:cut_shape].shape)],
y_entity_test],
callbacks=[early_stopping])