-
Notifications
You must be signed in to change notification settings - Fork 0
/
config.py
607 lines (539 loc) · 26.7 KB
/
config.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
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
# -*- coding: utf-8 -*-
# @Author : William
# @Project : TextGAN-william
# @FileName : config.py
# @Time : Created at 2019-03-18
# @Blog : http://zhiweil.ml/
# @Description :
# Copyrights (C) 2018. All Rights Reserved.
import time
from time import strftime, localtime
import os
import re
import torch
# ===Program===
if_test = False
if_eval = False
if_gen = False
if_robust = False
if_monit = False
if_old_version = False
if_eval_score = True
if_flex_com = False
if_train_defend = False
if_predict = False
if_train_val = False
if_eval_robust = False
if_show_sentence = False
if_accept_nocomment = True
if_hotflip_retrain = False
min_topic = {}
max_topic = {}
min_topic['gossipcopWithContent20x'] = 7
max_topic['gossipcopWithContent20x'] = 16
min_topic['rumors20x'] = 9
max_topic['rumors20x'] = 12
flg_eval_g = False
flg_eval_pre = False
flg_eval_copycat = False
flg_eval_hotflip = False
flg_eval_unitrigger = False
trigger_length = 3
trigger_beam_size = 2
trigger_neighbor = 30
if_trigger_retrain = False
if_textbugger_retrain = False
flg_eval_textbugger = False
flg_eval_baseline = False
attack_ratio = 1.0
eval_outfile = "Quality"
eval_num = 1
load_gen_adv_path = None
load_dis_adv_path = None
eval_topic_num = 1
atk_num_comment = 1
start_index = {}
start_index['gossipcopWithContent20x'] = 1
start_index['rumors20x'] = 0
CUDA = True
if_save = True
data_shuffle = False # False
oracle_pretrain = True # True
# gen_pretrain = True
dis_pretrain = False
clas_pretrain = False
visdom_ip = "207.148.26.49"
eval_g_path = ""
eval_style_path = ""
eval_pre_attack_path = ""
attack_mode = "target_real"
user = "DefaultUser"
blackbox_using_path = ""
run_model = 'relgan' # seqgan, leakgan, maligan, jsdgan, relgan, sentigan
k_label = 2 # num of labels, >=2
gen_init = 'truncated_normal' # normal, uniform, truncated_normal
dis_init = 'truncated_normal' # normal, uniform, truncated_normal
# ===Oracle or Real, type===
if_real_data = True # if use real data
dataset = 'gossipcopWithContent20x' # oracle, image_coco, emnlp_news, amazon_app_book, mr15
model_type = 'RMC' # vanilla, RMC (custom)
f_type = 'LR' #can be LR, CNN, RNN
h_type = 'CNN' #can be LR, CNN, RNN
loss_type = 'rsgan' # standard, JS, KL, hinge, tv, LS, rsgan (for RelGAN)
vocab_size = 10000 # oracle: 5000, coco: 6613, emnlp: 5255, amazon_app_book: 6418, mr15: 6289
max_seq_len = 26 # oracle: 20, coco: 37, emnlp: 51, amazon_app_book: 40
ADV_train_epoch = 5000 # SeqGAN, LeakGAN-200, RelGAN-3000
extend_vocab_size = 0 # plus test data, only used for Classifier
temp_adpt = 'exp' # no, lin, exp, log, sigmoid, quad, sqrt
temperature = 1000 # largest temperature
# ===Special===
sample_limit = 10
F_clf_lr = 0.001
F_train_epoch = 50
F_feature_dim = 45
f_embed_dim = 16
f_patience_epoch = 3
f_batch_size = 64
f_dropout_comment = 0.5
f_dropout_content = 0.95
rnn_hidden_dim = 32
num_topics = 14
H_clf_lr = 0.001
H_train_epoch = 50
h_embed_dim = 16
h_patience_epoch = 2
h_batch_size = 64
t_batch_size = 16
gen_adv_f_lr = 1e-4
g_f_step = 1
g_f_alpha = 1.0 # alpha multiplier for loss backprobagation from F to optimize G
if_adapt_alpha = False
gen_adv_h_lr = 5e-4
g_h_step = 1
g_h_alpha = 0.0
max_num_data = 50000
min_input_sent_length = 3
max_num_comment = 10
max_num_comment_test = 1
max_test_comment = 2
unk_threshold = 2
spelling_thres = 1
coherency_thres = 0.1
# max_content_len = 150
repeat_gen_pretrain = int(True)
load_gen_pretrain = int(True)
train_with_content = True
flg_train_f = False
flg_train_h = False
flg_adv_f = False
flg_adv_h = False
if_pretrain_embedding = False
random_num_article = 3
# ===Basic Train===
samples_num = 10000 # 10000, mr15: 2000,
samples_num2 = 128
MLE_train_epoch = 150 # SeqGAN-80, LeakGAN-8, RelGAN-150
PRE_clas_epoch = 10
inter_epoch = 15 # LeakGAN-10
batch_size = 256 # 64
mle_batch_size = 256
start_letter = 1
padding_idx = 0
unk_idx = 2
start_token = 'BOS'
end_token = 'EOS'
padding_token = '_'
unk_token = '<unk>'
gen_lr = 0.01 # 0.01
gen_adv_lr = 1e-4 # RelGAN-1e-4
dis_lr = 1e-4 # SeqGAN,LeakGAN-1e-2, RelGAN-1e-4
clas_lr = 1e-3
clip_norm = 5.0
pre_log_step = 50
adv_log_step = 100
train_data = 'dataset/' + dataset + '.txt'
test_data = 'dataset/testdata/' + dataset + '_test.txt'
cat_train_data = 'dataset/' + dataset + '_cat{}.txt'
cat_test_data = 'dataset/testdata/' + dataset + '_cat{}_test.txt'
content_embeddings = 'dataset/' + dataset + '_content_embeddings.npz'
comment_embeddings = 'dataset/' + dataset + '_comment_embeddings.npz'
h_train_data_file = 'dataset/' + dataset + '_coherence_train_data.txt'
h_train_label_file = 'dataset/' + dataset + '_coherence_train_label.txt'
h_dev_data_file = 'dataset/' + dataset + '_coherence_val_data.txt'
h_dev_label_file = 'dataset/' + dataset + '_coherence_val_label.txt'
h_test_data_file = 'dataset/' + dataset + '_coherence_test_data.txt'
h_test_label_file = 'dataset/' + dataset + '_coherence_test_label.txt'
f_train_data_file = 'dataset/' + dataset + '_detect_train_data.txt'
f_train_label_file = 'dataset/' + dataset + '_detect_train_label.txt'
f_train_val_data_file = 'dataset/' + dataset + '_detect_train_val_data.txt'
f_train_val_label_file = 'dataset/' + dataset + '_detect_train_val_label.txt'
f_dev_data_file = 'dataset/' + dataset + '_detect_val_data.txt'
f_dev_label_file = 'dataset/' + dataset + '_detect_val_label.txt'
f_test_data_file = 'dataset/' + dataset + '_detect_test_data.txt'
f_test_label_file = 'dataset/' + dataset + '_detect_test_label.txt'
topic_file = 'dataset/' + dataset + '_topic.pkl'
# ===Metrics===
use_nll_oracle = False
use_nll_gen = True
use_nll_div = False
use_bleu = True
use_self_bleu = False
use_clas_acc = False
use_ppl = False
# ===Generator===
ADV_g_step = 1 # 1
rollout_num = 16 # 4
gen_embed_dim = 32 # 32
gen_hidden_dim = 32 # 32
goal_size = 16 # LeakGAN-16
step_size = 4 # LeakGAN-4
mem_slots = 1 # RelGAN-1
num_heads = 2 # RelGAN-2
head_size = 256 # RelGAN-256 #changed from 64 # needs to change content_feature_dim
content_feature_dim = head_size #need to change head_size
mmd_mean = 2
mmd_kernel_num = 2
topic_epochs = 500
topic_lr = 0.0001
topic_gen_lr = 0.0001
ADV_topic_step = 1
flg_adv_t = True
flg_adv_d = True
# ===Discriminator===
d_step = 5 # SeqGAN-50, LeakGAN-5
d_epoch = 3 # SeqGAN,LeakGAN-3
ADV_d_step = 5 # SeqGAN,LeakGAN,RelGAN-5
ADV_d_epoch = 3 # SeqGAN,LeakGAN-3
dis_embed_dim = 64 #changed
dis_hidden_dim = 64 #changed
num_rep = 64 # RelGAN
device = 0
# ===log===
log_time_str = strftime("%m%d_%H%M_%S", localtime())
log_filename = strftime("log/log_%s" % log_time_str)
if os.path.exists(log_filename + '.txt'):
i = 2
while True:
if not os.path.exists(log_filename + '_%d' % i + '.txt'):
log_filename = log_filename + '_%d' % i
break
i += 1
log_filename = log_filename + '.txt'
# # Automatically choose GPU or CPU
# if torch.cuda.is_available() and torch.cuda.device_count() > 0:
# os.system('nvidia-smi -q -d Utilization > gpu')
# with open('gpu', 'r') as _tmpfile:
# util_gpu = list(map(int, re.findall(r'Gpu\s+:\s*(\d+)\s*%', _tmpfile.read())))
# os.remove('gpu')
# if len(util_gpu):
# device = util_gpu.index(min(util_gpu))
# else:
# device = 0
# else:
# device = -1
# # device=1
# # print('device: ', device)
# torch.cuda.set_device(device)
# ===Save Model and samples===
root_folder = "./"
save_root = root_folder + '/save/{}/{}/{}_{}_lt-{}_sl{}_temp{}_T{}/'.format(time.strftime("%Y%m%d"),
dataset, run_model, model_type,
loss_type, max_seq_len,
temperature,
log_time_str)
save_samples_root = save_root + 'samples/'
save_model_root = save_root + 'models/'
oracle_state_dict_path = 'pretrain/oracle_data/oracle_lstm.pt'
oracle_samples_path = 'pretrain/oracle_data/oracle_lstm_samples_{}.pt'
multi_oracle_state_dict_path = 'pretrain/oracle_data/oracle{}_lstm.pt'
multi_oracle_samples_path = 'pretrain/oracle_data/oracle{}_lstm_samples_{}.pt'
pretrain_root = root_folder+'/pretrain/{}/'.format(dataset if if_real_data else 'oracle_data')
pretrained_gen_path = pretrain_root + 'gen_MLE_pretrain_{}_{}_{}_sl{}_sn{}_hs{}_maxsn{}_vs{}_{}cmt.pt'.format(dataset, run_model, model_type,
max_seq_len, samples_num, head_size, max_num_data, vocab_size, max_num_comment)
pretrained_f_path = pretrain_root + 'F_pretrain_{}_{}_sl{}_sn{}_vs{}_{}cmt.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size, max_num_comment)
pretrained_f_path_CNN = pretrain_root + 'F_pretrain_{}_{}_sl{}_sn{}_vs{}_{}cmt_CNN.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size, max_num_comment)
pretrained_f_path_RNN = pretrain_root + 'F_pretrain_{}_{}_sl{}_sn{}_vs{}_{}cmt_Glove{}_RNN.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size, max_num_comment, f_embed_dim if if_pretrain_embedding else False)
pretrained_h_path = pretrain_root + 'H_pretrain_{}_{}_sl{}_sn{}_vs{}.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size)
pretrained_dis_path = pretrain_root + 'dis_pretrain_{}_{}_sl{}_sn{}_vs{}.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size)
pretrained_clas_path = pretrain_root + 'clas_pretrain_{}_{}_sl{}_sn{}_vs{}.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size)
topic_encoder_path = pretrain_root + 'topic_encoder_pretrained_MMD{}-{}'.format(mmd_mean, mmd_kernel_num)
signal_file = 'run_signal.txt'
tips = ''
# Init settings according to parser
def init_param(opt):
global run_model, model_type, loss_type, CUDA, device, data_shuffle, samples_num, vocab_size, \
MLE_train_epoch, ADV_train_epoch, inter_epoch, mle_batch_size, batch_size, max_seq_len, start_letter, padding_idx, \
gen_lr, gen_adv_lr, dis_lr, clip_norm, pre_log_step, adv_log_step, train_data, test_data, temp_adpt, \
temperature, oracle_pretrain, load_gen_pretrain, repeat_gen_pretrain, dis_pretrain, ADV_g_step, rollout_num, gen_embed_dim, \
gen_hidden_dim, goal_size, step_size, mem_slots, num_heads, head_size, d_step, d_epoch, \
ADV_d_step, ADV_d_epoch, dis_embed_dim, dis_hidden_dim, num_rep, log_filename, save_root, \
signal_file, tips, save_samples_root, save_model_root, if_real_data, pretrained_gen_path, \
pretrained_dis_path, pretrain_root, if_test, dataset, PRE_clas_epoch, oracle_samples_path, \
pretrained_clas_path, gen_init, dis_init, multi_oracle_samples_path, k_label, cat_train_data, cat_test_data, \
use_nll_oracle, use_nll_gen, use_nll_div, use_bleu, use_self_bleu, use_clas_acc, use_ppl, \
max_num_data, flg_train_f, flg_train_h, flg_adv_f, flg_adv_h, f_batch_size, h_batch_size, \
g_f_step, g_f_alpha, train_with_content, if_eval, eval_g_path, attack_mode, if_monit, \
h_train_data_file, h_train_label_file, h_dev_data_file, h_dev_label_file, h_test_data_file, h_test_label_file, \
f_train_data_file, f_train_label_file, f_dev_data_file, f_dev_label_file, content_embeddings, comment_embeddings, \
pretrained_f_path, pretrained_h_path, user, max_num_comment, if_eval_score, if_adapt_alpha,\
g_h_step, g_h_alpha, f_type, max_test_comment, content_feature_dim, if_pretrain_embedding, f_embed_dim, \
f_dropout_comment, f_dropout_content, pretrained_f_path_RNN, if_old_version, eval_num, eval_outfile, \
if_train_defend, if_predict, f_test_data_file, f_test_label_file, \
mmd_mean, mmd_kernel_num, topic_epochs, topic_lr, topic_gen_lr, ADV_topic_step, flg_adv_t, flg_adv_d, \
topic_file, samples_num2, num_topics, t_batch_size, if_save, max_num_comment_test, F_train_epoch, \
if_train_val, f_train_val_data_file, f_train_val_label_file, f_patience_epoch, root_folder, eval_topic_num, \
atk_num_comment, blackbox_using_path, flg_eval_g, flg_eval_pre, flg_eval_hotflip, flg_eval_copycat, \
if_show_sentence, random_num_article, flg_eval_unitrigger, trigger_length, trigger_beam_size, \
trigger_neighbor, if_trigger_retrain, flg_eval_textbugger, if_textbugger_retrain, attack_ratio, \
eval_style_path, if_gen, if_robust, eval_pre_attack_path, \
spelling_thres, coherency_thres, if_eval_robust, F_clf_lr, gen_adv_f_lr, flg_eval_baseline, if_hotflip_retrain, F_feature_dim
if_test = True if opt.if_test == 1 else False
if_eval = True if opt.if_eval == 1 else False
if_monit = True if opt.if_monit == 1 else False
if_predict = True if opt.if_predict == 1 else False
if_eval_score = True if opt.if_eval_score == 1 else False
if_train_defend = True if opt.if_train_defend ==1 else False
if_train_val = True if opt.if_train_val == 1 else False
flg_adv_d = True if opt.flg_adv_d == 1 else False
flg_eval_g = True if opt.flg_eval_g == 1 else False
flg_eval_pre = True if opt.flg_eval_pre == 1 else False
flg_eval_hotflip = True if opt.flg_eval_hotflip == 1 else False
flg_eval_copycat = True if opt.flg_eval_copycat == 1 else False
if_show_sentence = True if opt.if_show_sentence == 1 else False
flg_eval_unitrigger = True if opt.flg_eval_unitrigger == 1 else False
if_trigger_retrain = True if opt.if_trigger_retrain == 1 else False
flg_eval_textbugger = True if opt.flg_eval_textbugger == 1 else False
if_textbugger_retrain = True if opt.if_textbugger_retrain == 1 else False
if_gen = True if opt.if_gen == 1 else False
if_robust = True if opt.if_robust == 1 else False
if_eval_robust = True if opt.if_eval_robust == 1 else False
flg_eval_baseline = True if opt.flg_eval_baseline == 1 else False
if_hotflip_retrain = True if opt.if_hotflip_retrain == 1 else False
trigger_length = opt.trigger_length
trigger_neighbor = opt.trigger_neighbor
F_clf_lr = opt.F_clf_lr
eval_g_path = opt.eval_g_path
eval_style_path = opt.eval_style_path
eval_pre_attack_path = opt.eval_pre_attack_path
run_model = opt.run_model
k_label = opt.k_label
dataset = opt.dataset
model_type = opt.model_type
loss_type = opt.loss_type
if_real_data = True if opt.if_real_data == 1 else False
if_old_version = True if opt.if_old_version == 1 else False
CUDA = True if opt.cuda == 1 else False
device = opt.device
data_shuffle = opt.shuffle
gen_init = opt.gen_init
dis_init = opt.dis_init
eval_num = opt.eval_num
eval_outfile = opt.eval_outfile
samples_num2 = opt.max_num_eval
num_topics = opt.num_topics
if_save = opt.if_save
max_num_comment_test = opt.max_num_comment_test
F_train_epoch = opt.F_train_epoch
f_patience_epoch = opt.F_patience_epoch
eval_topic_num = opt.eval_topic_num
atk_num_comment = opt.atk_num_comment
blackbox_using_path = opt.blackbox_using_path
random_num_article = opt.random_num_article
trigger_beam_size = opt.trigger_beam_size
attack_ratio = opt.attack_ratio
spelling_thres = opt.spelling_thres
coherency_thres = opt.coherency_thres
gen_adv_f_lr = opt.gen_adv_f_lr
F_feature_dim = opt.F_feature_dim
samples_num = opt.samples_num
vocab_size = opt.vocab_size
MLE_train_epoch = opt.mle_epoch
PRE_clas_epoch = opt.clas_pre_epoch
ADV_train_epoch = opt.adv_epoch
inter_epoch = opt.inter_epoch
batch_size = opt.batch_size
mle_batch_size = opt.mle_batch_size
max_seq_len = opt.max_seq_len
start_letter = opt.start_letter
padding_idx = opt.padding_idx
gen_lr = opt.gen_lr
gen_adv_lr = opt.gen_adv_lr
dis_lr = opt.dis_lr
clip_norm = opt.clip_norm
pre_log_step = opt.pre_log_step
adv_log_step = opt.adv_log_step
train_data = opt.train_data
test_data = opt.test_data
temp_adpt = opt.temp_adpt
temperature = opt.temperature
g_f_step = opt.g_f_step
g_f_alpha = opt.g_f_alpha
f_embed_dim = opt.f_embed_dim
g_h_step = opt.g_h_step
g_h_alpha = opt.g_h_alpha
t_batch_size = opt.t_batch_size
if_adapt_alpha = opt.if_adapt_alpha
attack_mode = opt.attack_mode
user = opt.user
max_num_comment = opt.max_num_comment
max_test_comment = opt.max_test_comment
f_type = opt.f_type
f_dropout_comment = opt.f_dropout_comment
f_dropout_content = opt.f_dropout_content
mmd_mean = opt.mmd_mean
mmd_kernel_num = opt.mmd_kernel_num
topic_epochs = opt.topic_epochs
topic_lr = opt.topic_lr
topic_gen_lr = opt.topic_gen_lr
ADV_topic_step = opt.adv_topic_step
oracle_pretrain = True if opt.ora_pretrain == 1 else False
load_gen_pretrain = True if opt.load_gen_pretrain == 1 else False
dis_pretrain = True if opt.dis_pretrain == 1 else False
repeat_gen_pretrain = True if opt.gen_pretrain == 1 else False
flg_adv_f = True if opt.flg_adv_f == 1 else False
flg_adv_h = True if opt.flg_adv_h == 1 else False
flg_train_f = True if opt.flg_train_f == 1 else False
flg_train_h = True if opt.flg_train_h == 1 else False
train_with_content = True if opt.train_with_content == 1 else False
if_pretrain_embedding = True if opt.if_pretrain_embedding == 1 else False
max_num_data = opt.max_num_data
ADV_g_step = opt.adv_g_step
rollout_num = opt.rollout_num
gen_embed_dim = opt.gen_embed_dim
gen_hidden_dim = opt.gen_hidden_dim
goal_size = opt.goal_size
step_size = opt.step_size
mem_slots = opt.mem_slots
num_heads = opt.num_heads
head_size = opt.head_size
content_feature_dim = head_size
f_batch_size = opt.f_batch_size
h_batch_size = opt.h_batch_size
d_step = opt.d_step
d_epoch = opt.d_epoch
ADV_d_step = opt.adv_d_step
ADV_d_epoch = opt.adv_d_epoch
dis_embed_dim = opt.dis_embed_dim
dis_hidden_dim = opt.dis_hidden_dim
num_rep = opt.num_rep
flg_adv_t = True if opt.flg_adv_t == 1 else False
use_nll_oracle = True if opt.use_nll_oracle == 1 else False
use_nll_gen = True if opt.use_nll_gen == 1 else False
use_nll_div = True if opt.use_nll_div == 1 else False
use_bleu = True if opt.use_bleu == 1 else False
use_self_bleu = True if opt.use_self_bleu == 1 else False
use_clas_acc = True if opt.use_clas_acc == 1 else False
use_ppl = True if opt.use_ppl == 1 else False
log_filename = opt.log_file
signal_file = opt.signal_file
tips = opt.tips
root_folder = opt.root_folder
# CUDA device
# torch.cuda.set_device(device)
# Save path
save_root = root_folder + '/save/{}/{}/{}_{}_lt-{}_sl{}_temp{}_T{}/'.format(time.strftime("%Y%m%d"),
dataset, run_model, model_type,
loss_type, max_seq_len,
temperature,
log_time_str)
save_samples_root = save_root + 'samples/'
save_model_root = save_root + 'models/'
train_data = 'dataset/' + dataset + '.txt'
test_data = 'dataset/testdata/' + dataset + '_test.txt'
train_data_content = 'dataset/' + dataset + '.txt'
test_data_content = 'dataset/testdata/' + dataset + '_test.txt'
cat_train_data = 'dataset/' + dataset + '_cat{}.txt'
cat_test_data = 'dataset/testdata/' + dataset + '_cat{}_test.txt'
# if max_seq_len == 40:
# oracle_samples_path = 'pretrain/oracle_data/oracle_lstm_samples_{}_sl40.pt'
# multi_oracle_samples_path = 'pretrain/oracle_data/oracle{}_lstm_samples_{}_sl40.pt'
pretrain_root = root_folder+'/pretrain/{}/'.format(dataset if if_real_data else 'oracle_data')
pretrained_gen_path = pretrain_root + 'gen_MLENEW_pretrain_{}_{}_{}_sl{}_sn{}_hs{}_maxsn{}_vs{}.pt'.format(dataset, run_model, model_type,
max_seq_len, samples_num, head_size, max_num_data, vocab_size)
pretrain_root = root_folder+'/pretrain/{}/'.format(dataset if if_real_data else 'oracle_data')
pretrained_f_path = pretrain_root + 'F_pretrain_{}_{}_sl{}_sn{}_vs{}_{}cmt_Glove{}.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size, max_num_comment, f_embed_dim if if_pretrain_embedding else False)
pretrained_f_path_CNN = pretrain_root + 'F_pretrain_{}_{}_sl{}_sn{}_vs{}_{}cmt_Glove{}_CNN.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size, max_num_comment, f_embed_dim if if_pretrain_embedding else False)
pretrained_f_path_RNN = pretrain_root + 'F_pretrain_{}_{}_sl{}_sn{}_vs{}_{}cmt_Glove{}_RNN.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size, max_num_comment, f_embed_dim if if_pretrain_embedding else False)
pretrained_f_path_RNN2 = pretrain_root + 'F_pretrain_{}_{}_sl{}_sn{}_vs{}_{}cmt_Glove{}_RNN2.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size, max_num_comment, f_embed_dim if if_pretrain_embedding else False)
pretrained_f_path_CSI = pretrain_root + 'F_pretrain_{}_{}_sl{}_sn{}_vs{}_{}cmt_Glove{}_CSI.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size, max_num_comment, f_embed_dim if if_pretrain_embedding else False)
pretrained_f_path_TEXTCNN = pretrain_root + 'F_pretrain_{}_{}_sl{}_sn{}_vs{}_{}cmt_Glove{}_TEXTCNN.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size, max_num_comment, f_embed_dim if if_pretrain_embedding else False)
pretrained_f_path_TEXTCNN2 = pretrain_root + 'F_pretrain_{}_{}_sl{}_sn{}_vs{}_{}cmt_Glove{}_TEXTCNN2.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size, max_num_comment, f_embed_dim if if_pretrain_embedding else False)
pretrained_f_path_TEXTCNN3 = pretrain_root + 'F_pretrain_{}_{}_sl{}_sn{}_vs{}_{}cmt_Glove{}_TEXTCNN3.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size, max_num_comment, f_embed_dim if if_pretrain_embedding else False)
pretrained_h_path = pretrain_root + 'H_pretrain_{}_{}_sl{}_sn{}_vs{}.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size)
pretrained_dis_path = pretrain_root + 'dis_pretrain_{}_{}_sl{}_sn{}_vs{}.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size)
pretrained_clas_path = pretrain_root + 'clas_pretrain_{}_{}_sl{}_sn{}_vs{}.pt'.format(run_model, model_type, max_seq_len,
samples_num, vocab_size)
topic_encoder_path = pretrain_root + 'topic_encoder_pretrained_MMD{}-{}'.format(mmd_mean, mmd_kernel_num)
if f_type == "CNN":
pretrained_f_path = pretrained_f_path_CNN
elif f_type == "RNN":
pretrained_f_path = pretrained_f_path_RNN
elif f_type == "RNN2":
pretrained_f_path = pretrained_f_path_RNN2
elif f_type == "CSI":
pretrained_f_path = pretrained_f_path_CSI
elif f_type == "TEXTCNN":
pretrained_f_path = pretrained_f_path_TEXTCNN
elif f_type == "TEXTCNN2":
pretrained_f_path = pretrained_f_path_TEXTCNN2
elif f_type == "TEXTCNN3":
pretrained_f_path = pretrained_f_path_TEXTCNN3
content_embeddings = 'dataset/' + dataset + '_content_embeddings.npz'
comment_embeddings = 'dataset/' + dataset + '_comment_embeddings.npz'
h_train_data_file = 'dataset/' + dataset + '_coherence_train_data.txt'
h_train_label_file = 'dataset/' + dataset + '_coherence_train_label.txt'
h_dev_data_file = 'dataset/' + dataset + '_coherence_val_data.txt'
h_dev_label_file = 'dataset/' + dataset + '_coherence_val_label.txt'
h_test_data_file = 'dataset/' + dataset + '_coherence_test_data.txt'
h_test_label_file = 'dataset/' + dataset + '_coherence_test_label.txt'
f_train_data_file = 'dataset/' + dataset + '_detect_train_data.txt'
f_train_label_file = 'dataset/' + dataset + '_detect_train_label.txt'
f_train_val_data_file = 'dataset/' + dataset + '_detect_train_val_data.txt'
f_train_val_label_file = 'dataset/' + dataset + '_detect_train_val_label.txt'
f_dev_data_file = 'dataset/' + dataset + '_detect_val_data.txt'
f_dev_label_file = 'dataset/' + dataset + '_detect_val_label.txt'
f_test_data_file = 'dataset/' + dataset + '_detect_test_data.txt'
f_test_label_file = 'dataset/' + dataset + '_detect_test_label.txt'
topic_file = 'dataset/' + dataset + '_topic_{}.pkl'.format(num_topics)
pretrained_f_path = opt.attack_f_path if opt.attack_f_path != "" else pretrained_f_path
pretrained_h_path = opt.attack_h_path if opt.attack_h_path != "" else pretrained_h_path
pretrained_gen_path = opt.pretrained_gen_path if opt.pretrained_gen_path != "" else pretrained_gen_path
if if_train_val:
f_train_data_file = f_train_val_data_file
f_train_label_file = f_train_val_label_file
# Assertion
assert k_label >= 2, 'Error: k_label = {}, which should be >=2!'.format(k_label)
# Create Directory
dir_list = ['save', 'savefig', 'log', 'pretrain', 'dataset',
'pretrain/{}'.format(dataset if if_real_data else 'oracle_data')]
if not if_test:
dir_list.extend([save_root, save_samples_root, save_model_root])
for d in dir_list:
if not os.path.exists(d):
os.makedirs(d)
# Assertion
assert k_label >= 2, 'Error: k_label = {}, which should be >=2!'.format(k_label)
# Create Directory
dir_list = ['save', 'savefig', 'log', 'pretrain', 'dataset',
'pretrain/{}'.format(dataset if if_real_data else 'oracle_data')]
if not if_test:
dir_list.extend([save_root, save_samples_root, save_model_root])
for d in dir_list:
if not os.path.exists(d):
os.makedirs(d)