-
Notifications
You must be signed in to change notification settings - Fork 15
/
train.py
202 lines (150 loc) · 6.52 KB
/
train.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
196
197
198
199
200
201
# coding: utf-8
import os
import time
import sys
import yaml
import numpy as np
import pandas as pd
from src.util import ExeDataset,write_pred
from src.model import MalConv
from torch.utils.data import DataLoader
import torch
import torch.nn as nn
import torch.optim as optim
from torch.autograd import Variable
# 加载配置文件
try:
config_path = "/Users/apple/Documents/GitHub/Deep learning for malware detection/MalConv-Pytorch/config/example.yaml" #sys.argv[1]
seed = 10 #int(sys.argv[2])
conf = yaml.load(open(config_path,'r'))
except:
print('Usage: python3 run_exp.py <config file path> <seed>')
sys.exit()
# 开始试验
exp_name = conf['exp_name']+'_sd_'+str(seed)
print('Experiment:')
print('\t',exp_name)
np.random.seed(seed)
torch.manual_seed(seed)
train_data_path = conf['train_data_path']
train_label_path = conf['train_label_path']
valid_data_path = conf['valid_data_path']
valid_label_path = conf['valid_label_path']
log_dir = conf['log_dir']
pred_dir = conf['pred_dir']
checkpoint_dir = conf['checkpoint_dir']
log_file_path = log_dir+exp_name+'.log'
chkpt_acc_path = checkpoint_dir+exp_name+'.model'
pred_path = pred_dir+exp_name+'.pred'
# Parameters
use_gpu = conf['use_gpu']
use_cpu = conf['use_cpu']
learning_rate = conf['learning_rate']
max_step = conf['max_step']
test_step = conf['test_step']
batch_size = conf['batch_size']
first_n_byte = conf['first_n_byte']
window_size = conf['window_size']
display_step = conf['display_step']
sample_cnt = conf['sample_cnt']
# Load Ground Truth.
tr_label_table = pd.read_csv(train_label_path,header=None,index_col=0)
tr_label_table.index=tr_label_table.index.str.upper()
tr_label_table = tr_label_table.rename(columns={1:'ground_truth'})
val_label_table = pd.read_csv(valid_label_path,header=None,index_col=0)
val_label_table.index=val_label_table.index.str.upper()
val_label_table = val_label_table.rename(columns={1:'ground_truth'})
# Merge Tables and remove duplicate
tr_table = tr_label_table.groupby(level=0).last()
del tr_label_table
val_table = val_label_table.groupby(level=0).last()
del val_label_table
tr_table = tr_table.drop(val_table.index.join(tr_table.index, how='inner'))
print('Training Set:')
print('\tTotal',len(tr_table),'files')
print('\tMalware Count :',tr_table['ground_truth'].value_counts()[1])
print('\tGoodware Count:',tr_table['ground_truth'].value_counts()[0])
print('Validation Set:')
print('\tTotal',len(val_table),'files')
print('\tMalware Count :',val_table['ground_truth'].value_counts()[1])
print('\tGoodware Count:',val_table['ground_truth'].value_counts()[0])
if sample_cnt != 1:
tr_table = tr_table.sample(n=sample_cnt,random_state=seed)
dataset_test = ExeDataset(list(tr_table.index), train_data_path, list(tr_table.ground_truth),first_n_byte)
dataloader = DataLoader(dataset_test , batch_size=batch_size, shuffle=True, num_workers=use_cpu)
validloader = DataLoader(ExeDataset(list(val_table.index), valid_data_path, list(val_table.ground_truth),first_n_byte),
batch_size=batch_size, shuffle=False, num_workers=use_cpu)
valid_idx = list(val_table.index)
del tr_table
del val_table
malconv = MalConv(input_length=first_n_byte,window_size=window_size)
bce_loss = nn.BCEWithLogitsLoss()
adam_optim = optim.Adam([{'params':malconv.parameters()}],lr=learning_rate)
sigmoid = nn.Sigmoid()
if use_gpu:
malconv = malconv.cuda()
bce_loss = bce_loss.cuda()
sigmoid = sigmoid.cuda()
step_msg = 'step-{}-loss-{:.6f}-acc-{:.4f}-time-{:.2f}'
valid_msg = 'step-{}-tr_loss-{:.6f}-tr_acc-{:.4f}-val_loss-{:.6f}-val_acc-{:.4f}'
log_msg = '{}, {:.6f}, {:.4f}, {:.6f}, {:.4f}, {:.2f}'
history = {}
history['tr_loss'] = []
history['tr_acc'] = []
log = open(log_file_path,'w')
log.write('step,tr_loss, tr_acc, val_loss, val_acc, time\n')
valid_best_acc = 0.0
total_step = 0
step_cost_time = 0
while total_step < max_step:
# Training
for step,batch_data in enumerate(dataloader):
start = time.time()
adam_optim.zero_grad()
cur_batch_size = batch_data[0].size(0)
exe_input = batch_data[0].cuda() if use_gpu else batch_data[0]
exe_input = Variable(exe_input.long(),requires_grad=False)
label = batch_data[1].cuda() if use_gpu else batch_data[1]
label = Variable(label.float(),requires_grad=False)
pred = malconv(exe_input)
loss = bce_loss(pred,label)
loss.backward()
adam_optim.step()
history['tr_loss'].append(loss.cpu().data.numpy())
history['tr_acc'].extend(list(label.cpu().data.numpy().astype(int)==(sigmoid(pred).cpu().data.numpy()+0.5).astype(int)))
step_cost_time = time.time()-start
if (step+1)%display_step == 0:
print(step_msg.format(total_step,np.mean(history['tr_loss']),
np.mean(history['tr_acc']),step_cost_time),end='\r',flush=True)
total_step += 1
# Interupt for validation
if total_step%test_step ==0:
break
# Testing
history['val_loss'] = []
history['val_acc'] = []
history['val_pred'] = []
for _,val_batch_data in enumerate(validloader):
cur_batch_size = val_batch_data[0].size(0)
exe_input = val_batch_data[0].cuda() if use_gpu else val_batch_data[0]
exe_input = Variable(exe_input.long(),requires_grad=False)
label = val_batch_data[1].cuda() if use_gpu else val_batch_data[1]
label = Variable(label.float(),requires_grad=False)
pred = malconv(exe_input)
loss = bce_loss(pred,label)
history['val_loss'].append(loss.cpu().data.numpy())
history['val_acc'].extend(list(label.cpu().data.numpy().astype(int)==(sigmoid(pred).cpu().data.numpy()+0.5).astype(int)))
history['val_pred'].append(list(sigmoid(pred).cpu().data.numpy()))
print(log_msg.format(total_step, np.mean(history['tr_loss']), np.mean(history['tr_acc']),
np.mean(history['val_loss']), np.mean(history['val_acc']),step_cost_time),
file=log,flush=True)
print(valid_msg.format(total_step,np.mean(history['tr_loss']),np.mean(history['tr_acc']),
np.mean(history['val_loss']),np.mean(history['val_acc'])))
if valid_best_acc < np.mean(history['val_acc']):
valid_best_acc = np.mean(history['val_acc'])
torch.save(malconv,chkpt_acc_path)
print('Checkpoint saved at',chkpt_acc_path)
write_pred(history['val_pred'],valid_idx,pred_path)
print('Prediction saved at', pred_path)
history['tr_loss'] = []
history['tr_acc'] = []