Skip to content
Navigation Menu
{{ message }}
-
Notifications
You must be signed in to change notification settings - Fork 27
Expand file tree
/
Copy pathstringc2.py
More file actions
383 lines (320 loc) · 13 KB
/
Copy pathstringc2.py
File metadata and controls
383 lines (320 loc) · 13 KB
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
#
# Copyright (c) 2020 Adobe Systems Incorporated. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
import torch
import torch.nn as nn
import optparse
import sys
import json
import numpy as np
import random
import tqdm
class Encodings:
def __init__(self, filename=None):
self._char2int = {'<PAD>': 0, '<UNK>': 1}
self._label2int = {'<PAD>': 0}
self._label_list = ['<PAD>']
if filename is not None:
self.load(filename)
def save(self, filename):
json.dump({'char2int': self._char2int, 'label2int': self._label2int},
open(filename, 'w'))
def load(self, file):
if isinstance(file, str):
stream = open(file, 'r')
else:
stream = file
obj = json.load(stream)
self._char2int = obj['char2int']
self._label2int = obj['label2int']
self._label_list = [None for _ in range(len(self._label2int))]
for t in self._label2int:
self._label_list[self._label2int[t]] = t
def update_encodings(self, dataset, cutoff=2):
char2count = {}
for entry in tqdm.tqdm(dataset):
text = entry[0]
label = entry[1]
for char in text:
char = char.lower()
if char in char2count:
char2count[char] += 1
else:
char2count[char] = 1
for ttype in label:
if ttype not in self._label2int:
self._label2int[ttype] = len(self._label2int)
self._label_list.append(ttype)
for char in char2count:
if char not in self._char2int and char2count[char] > cutoff:
self._char2int[char] = len(self._char2int)
class CTaggerConfig:
def __init__(self):
self.char_emb_size = 100
self.rnn_layers = 2
self.rnn_size = 100
self.hidden = 500
def save(self, filename):
json.dump({'char_emb_size': self.char_emb_size, 'rnn_layers': self.rnn_layers, 'rnn_size': self.rnn_size,
'hidden': self.hidden},
open(filename, 'w'))
def load(self, file):
if isinstance(file, str):
stream = open(file, 'r')
else:
stream = file
obj = json.load(stream)
self.char_emb_size = obj['char_emb_size']
self.rnn_size = obj['rnn_size']
self.rnn_layers = obj['rnn_layers']
self.hidden = obj['hidden']
class CTagger(nn.Module):
def __init__(self, config, encodings):
super(CTagger, self).__init__()
self._config = config
self._encodings = encodings
self._char_emb = nn.Embedding(len(encodings._char2int), config.char_emb_size, padding_idx=0)
self._case_emb = nn.Embedding(4, 16, padding_idx=0)
self._rnn = nn.LSTM(config.char_emb_size + 16, config.rnn_size, config.rnn_layers, batch_first=True,
bidirectional=True)
self._hidden = nn.Sequential(nn.Linear(config.rnn_size * 2, config.hidden), nn.Tanh(), nn.Dropout(0.5))
self._softmax_type = nn.Linear(config.hidden, len(encodings._label2int))
def _make_input(self, word_list):
# we pad domain names and feed them in reversed character order to the LSTM
max_seq_len = max([len(word) for word in word_list])
x_char = np.zeros((len(word_list), max_seq_len))
x_case = np.zeros((len(word_list), max_seq_len))
for iBatch in range(x_char.shape[0]):
word = word_list[iBatch]
for index in range(len(word)):
char = word[index]
case_idx = 0
if char.lower() == char.upper():
case_idx = 1 # symbol
elif char.lower() != char:
case_idx = 2 # uppercase
else:
case_idx = 3 # lowercase
char = char.lower()
if char in self._encodings._char2int:
char_idx = self._encodings._char2int[char]
else:
char_idx = 1 # UNK
x_char[iBatch, index] = char_idx
x_case[iBatch, index] = case_idx
return x_char, x_case
def forward(self, string_list):
x_char, x_case = self._make_input(string_list)
x_char = torch.tensor(x_char, dtype=torch.long, device=self._get_device())
x_case = torch.tensor(x_case, dtype=torch.long, device=self._get_device())
hidden = torch.cat([self._char_emb(x_char), self._case_emb(x_case)], dim=-1)
hidden = torch.dropout(hidden, 0.5, self.training)
output, _ = self._rnn(hidden)
hidden = self._hidden(output)
return self._softmax_type(hidden)
def save(self, path):
torch.save(self.state_dict(), path)
def load(self, path):
self.load_state_dict(torch.load(path, map_location='cpu'))
def _get_device(self):
if self._char_emb.weight.device.type == 'cpu':
return 'cpu'
return '{0}:{1}'.format(self._char_emb.weight.device.type, str(self._char_emb.weight.device.index))
def _load_dataset(filename):
lines = open(filename, encoding='utf-8').readlines()
dataset = []
for ii in range(len(lines) // 2):
string = lines[ii * 2][:-1]
mask = lines[ii * 2 + 1][:-1]
dataset.append((string, mask))
return dataset
def _eval(model, dataset, encodings):
model.eval()
test_x, test_y = _make_batches(dataset, batch_size=128)
total_t = 0
ok_t = 0
with torch.no_grad():
pgb = tqdm.tqdm(zip(test_x, test_y), total=len(test_x), ncols=80, desc='\t\t\t\t')
for x, y in pgb:
y_pred_t = model(x)
y_tar_t = _get_targets(y, encodings).reshape(-1)
y_pred_t = torch.argmax(y_pred_t, dim=-1).detach().cpu().numpy().reshape(-1)
for y_t_t, y_p_t in zip(y_tar_t, y_pred_t):
if y_t_t != 0:
total_t += 1
if y_t_t == y_p_t:
ok_t += 1
return ok_t / total_t
def _make_batches(dataset, batch_size=32):
batches_x = []
batches_y = []
batch_x = []
batch_y = []
for entry in dataset:
domain = entry[0]
t = entry[1]
batch_x.append(domain)
batch_y.append(t)
if len(batch_x) == batch_size:
batches_x.append(batch_x)
batches_y.append(batch_y)
batch_x = []
batch_y = []
if len(batch_x) != 0:
batches_x.append(batch_x)
batches_y.append(batch_y)
return batches_x, batches_y
def _get_targets(y, encodings):
max_len = max([len(yy) for yy in y])
y_t = np.zeros((len(y), max_len), dtype=np.long)
for i in range(len(y)):
for j in range(max_len):
if j < len(y[i]):
y_t[i, j] = encodings._label2int[y[i][j]]
return y_t
def _generate_dataset(count):
from training import generate_next_cmd
dataset = []
for ii in range(count):
cmd, mask = generate_next_cmd()
dataset.append((cmd, mask))
return dataset
def _start_train(params):
eval_at = 5000
if params.resume:
encodings = Encodings('{0}.encodings'.format(params.output_base))
else:
sys.stdout.write('Generating new random data...')
sys.stdout.flush()
trainset = _generate_dataset(int(eval_at * 4 * params.batch_size))
sys.stdout.write('done\n')
encodings = Encodings()
encodings.update_encodings(trainset)
print('chars={0}, types={1}'.format(len(encodings._char2int), len(encodings._label2int)))
print(encodings._label2int)
config = CTaggerConfig()
if params.resume:
config.load('{0}.conf'.format(params.output_base))
model = CTagger(config, encodings)
model.to(params.device)
if params.resume:
model.load('{0}.last'.format(params.output_base))
optimizer = torch.optim.Adam(model.parameters())
criterion_t = torch.nn.CrossEntropyLoss(ignore_index=0)
patience_left = params.patience
best_type = 0 # _eval(model, devset, encodings)
encodings.save('{0}.encodings'.format(params.output_base))
config.save('{0}.conf'.format(params.output_base))
model.save('{0}.last'.format(params.output_base))
print("Deveset evaluation acc={0}".format(best_type))
epoch = 0
eval_at = 5000
while patience_left > 0:
sys.stdout.write('Generating new random data...')
sys.stdout.flush()
trainset = _generate_dataset(int(eval_at * params.batch_size))
devset = _generate_dataset(int(eval_at / 10 * params.batch_size))
sys.stdout.write('done\n')
sys.stdout.flush()
sys.stderr.flush()
epoch += 1
random.shuffle(trainset)
train_x, train_y = _make_batches(trainset, batch_size=params.batch_size)
sys.stdout.write('Starting epoch {0}\n'.format(epoch))
pgb = tqdm.tqdm(zip(train_x, train_y), total=len(train_x), ncols=80, desc='\tloss=N/A')
model.train()
total_loss = 0
cnt = 0
for x, y in pgb:
cnt += 1
if cnt % eval_at == 0:
patience_left -= 1
sys.stderr.flush()
sys.stderr.flush()
sys.stderr.write('\n\tEvaluating...')
sys.stderr.flush()
acc_t = _eval(model, devset, encodings)
sys.stderr.write(' acc={0}\n'.format(acc_t))
sys.stderr.flush()
filename = '{0}.last'.format(params.output_base)
sys.stderr.write('\t\tStoring {0}\n'.format(filename))
sys.stderr.flush()
model.save(filename)
if acc_t > best_type:
patience_left = params.patience
best_type = acc_t
filename = '{0}.bestType'.format(params.output_base)
sys.stderr.write('\t\tStoring {0}\n'.format(filename))
sys.stderr.flush()
model.save(filename)
sys.stderr.write('\n')
sys.stderr.flush()
model.train()
if patience_left <= 0:
print("Stopping with maximum patience reached")
sys.exit(0)
y_pred_t = model(x)
y_tar_t = _get_targets(y, encodings)
y_tar_t = torch.tensor(y_tar_t, dtype=torch.long, device=params.device)
y_pred = y_pred_t.view(-1, y_pred_t.shape[-1])
y_target = y_tar_t.view(-1)
if y_pred.shape[0] != y_target.shape[0]:
from ipdb import set_trace
set_trace()
loss = criterion_t(y_pred, y_target)
optimizer.zero_grad()
total_loss += loss.item()
pgb.set_description('\tloss={0:.4f}'.format(total_loss / cnt))
loss.backward()
optimizer.step()
sys.stdout.write('AVG train loss={0} \n'.format(total_loss / len(train_x)))
def _start_interactive(params):
encodings = Encodings('{0}.encodings'.format(params.output_base))
config = CTaggerConfig()
config.load('{0}.conf'.format(params.output_base))
model = CTagger(config, encodings)
model.load('{0}.bestType'.format(params.output_base))
model.to(params.device)
model.eval()
sys.stdout.write('>>> ')
sys.stdout.flush()
string = input()
while string != '/exit':
p_t = model([string])
p_d_t = torch.argmax(p_t, dim=-1).detach().cpu().numpy()
print("Results for \n{0}".format(string))
for ii in range(p_d_t.shape[-1]):
sys.stdout.write(encodings._label_list[p_d_t[0, ii]])
sys.stdout.write('\n')
print("")
sys.stdout.write('>>> ')
sys.stdout.flush()
string = input()
if __name__ == '__main__':
parser = optparse.OptionParser()
parser.add_option('--interactive', action='store_true', dest='interactive')
parser.add_option('--train', action='store_true', dest='train')
parser.add_option('--resume', action='store_true', dest='resume')
parser.add_option('--store', action='store', dest='output_base')
parser.add_option('--patience', action='store', dest='patience', type='int', default=20, help='(default=20)')
parser.add_option('--batch-size', action='store', dest='batch_size', default=32, type='int', help='(default=32)')
parser.add_option('--device', action='store', dest='device', default='cpu')
(params, _) = parser.parse_args(sys.argv)
if params.train:
_start_train(params)
elif params.interactive:
_start_interactive(params)
else:
parser.print_help()
You can’t perform that action at this time.