-
Notifications
You must be signed in to change notification settings - Fork 8
/
train.py
272 lines (195 loc) · 8.87 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
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
import os
import time
import numpy as np
import torch
import logging
from transformers import BertTokenizer
import tensorboard_logger as tb_logger
import arguments
from lib import evaluation
from lib import image_caption
from lib.vse import VSEModel
from lib.evaluation import i2t, t2i, AverageMeter, LogCollector, encode_data, compute_sim
from graph_lib import *
def main():
# Hyper Parameters
parser = arguments.get_argument_parser()
parser = extra_parameters(parser)
opt = parser.parse_args()
opt.model_name = opt.logger_name
# set the gpu-id for training
if not opt.multi_gpu:
torch.cuda.set_device(opt.gpu_id)
# create the folder for logger and checkpoint
if not os.path.exists(opt.model_name):
os.makedirs(opt.model_name)
# initialize logger
logging.basicConfig(filename=os.path.join(opt.logger_name, 'train.txt'), filemode='w',
format='%(asctime)s %(message)s', level=logging.INFO)
logger = logging.getLogger(__name__)
logger.info(opt)
tb_logger.configure(opt.logger_name, flush_secs=5)
# record parameters
arguments.save_parameters(opt, opt.logger_name)
# load tokenizer for TextEncoder
# tokenizer = BertTokenizer.from_pretrained(opt.bert_path)
tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
# get the train-set
train_loader = image_caption.get_train_loader(opt.data_path, tokenizer, opt.batch_size, opt.workers, opt)
# get the test-set
split = 'testall' if opt.dataset == 'coco' else 'test'
test_loader = image_caption.get_test_loader(opt.data_path, split, tokenizer, opt.batch_size, opt.workers, opt)
logger.info('Number of images for train-set: {}'.format(train_loader.dataset.num_images))
# load the multi-modal model
model = VSEModel(opt)
start_epoch = 0
# use the multi gpu
if (not model.is_data_parallel) and opt.multi_gpu:
model.make_data_parallel()
best_rsum = 0
# start the training process
for epoch in range(start_epoch, opt.num_epochs):
if epoch == 0:
logger.info('Log saving path: ' + opt.logger_name)
logger.info('Models saving path: ' + opt.model_name)
adjust_learning_rate(opt, model.optimizer, epoch)
# set hard negative for vse loss
if (epoch >= opt.vse_mean_warmup_epochs):
opt.max_violation = True
model.set_max_violation(opt.max_violation)
# train for one epoch
train(opt, train_loader, model, epoch)
# evaluate on test set for every epoch
rsum = validate(opt, test_loader, model)
# remember best rsum and save checkpoint
is_best = rsum > best_rsum
best_rsum = max(rsum, best_rsum)
logger.info("Epoch: [{}], Best rsum: {:.1f}".format(epoch, best_rsum))
# save the checkpoint
state = {'model': model.state_dict(), 'opt': opt, 'epoch': epoch + 1, 'best_rsum': best_rsum, 'Eiters': model.Eiters}
save_checkpoint(state, is_best, prefix=opt.model_name)
logger.info('Train finish.')
# evaluation after training process
logger.info('Evaluate the model')
base = opt.logger_name
logging.basicConfig(filename=os.path.join(base, 'eval.txt'), filemode='w',
format='%(asctime)s %(message)s', level=logging.INFO, force=True)
logger = logging.getLogger()
logger.info('Evaluating {}'.format(base))
model_path = os.path.join(base, 'model_best.pth')
# Save the final results for computing ensemble results
save_path = os.path.join(base, 'results_{}.npy'.format(opt.dataset)) if opt.save_results else None
if opt.dataset == 'coco':
# Evaluate COCO 5-fold 1K
# Evaluate COCO 5K
evaluation.evalrank(model_path, opt=opt, tokenizer=tokenizer, model=model, split='testall', fold5=True, save_path=save_path)
else:
# Evaluate Flickr30K
evaluation.evalrank(model_path, opt=opt, tokenizer=tokenizer, model=model, split='test', fold5=False, save_path=save_path)
logger.info('Evaluation finish')
def train(opt, train_loader, model, epoch):
logger = logging.getLogger(__name__)
batch_time = AverageMeter()
data_time = AverageMeter()
train_logger = LogCollector()
if epoch == 0:
logger.info('image encoder trainable parameters: {}M'.format(count_params(model.img_enc)))
logger.info('txt encoder trainable parameters: {}M'.format(count_params(model.txt_enc)))
logger.info('criterion trainable parameters: {}M'.format(count_params(model.criterion)))
end = time.time()
repeat_list = []
n_batch = len(train_loader.dataset) // opt.batch_size
model.train_start()
for i, train_data in enumerate(train_loader):
# measure data loading time
data_time.update(time.time() - end)
# make sure train logger is used
model.logger = train_logger
# Update the model
images, img_lengths, captions, lengths, ids, img_ids, repeat = train_data
model.train_emb(images, captions, lengths, image_lengths=img_lengths, img_ids=img_ids)
# measure elapsed time
batch_time.update(time.time() - end)
end = time.time()
if model.Eiters % opt.log_step == 0:
logging.info(
'Epoch: [{0}][{1}/{2}]\t'
'{e_log}\t'
'Batch-Time {batch_time.val:.2f} ({batch_time.avg:.2f})\t'
.format(
epoch, i+1, n_batch, batch_time=batch_time,
data_time=data_time, e_log=str(model.logger)))
# Record logs in tensorboard
tb_logger.log_value('epoch', epoch, step=model.Eiters)
tb_logger.log_value('step', i, step=model.Eiters)
tb_logger.log_value('batch_time', batch_time.val, step=model.Eiters)
tb_logger.log_value('data_time', data_time.val, step=model.Eiters)
model.logger.tb_log(tb_logger, step=model.Eiters)
return repeat_list
def validate(opt, val_loader, model):
logger = logging.getLogger(__name__)
model.val_start()
with torch.no_grad():
# compute the encoding for all the validation images and captions
img_embs, cap_embs = encode_data(model, val_loader, opt.log_step, logging.info)
# have repetitive image features
img_embs = img_embs[::5]
npts = img_embs.shape[0]
sims = compute_sim(img_embs, cap_embs)
(r1, r5, r10, medr, meanr) = i2t(npts, sims)
logging.info("Image to text (R@1, R@5, R@10): %.1f, %.1f, %.1f" % (r1, r5, r10))
(r1i, r5i, r10i, medri, meanr) = t2i(npts, sims)
logging.info("Text to image (R@1, R@5, R@10): %.1f, %.1f, %.1f" % (r1i, r5i, r10i))
# sum of recalls to be used for early stopping
currscore = r1 + r5 + r10 + r1i + r5i + r10i
logger.info('Current rsum is {}'.format(round(currscore, 1)))
# record metrics in tensorboard
tb_logger.log_value('r1', r1, step=model.Eiters)
tb_logger.log_value('r5', r5, step=model.Eiters)
tb_logger.log_value('r10', r10, step=model.Eiters)
tb_logger.log_value('medr', medr, step=model.Eiters)
tb_logger.log_value('meanr', meanr, step=model.Eiters)
tb_logger.log_value('r1i', r1i, step=model.Eiters)
tb_logger.log_value('r5i', r5i, step=model.Eiters)
tb_logger.log_value('r10i', r10i, step=model.Eiters)
tb_logger.log_value('medri', medri, step=model.Eiters)
tb_logger.log_value('meanr', meanr, step=model.Eiters)
tb_logger.log_value('rsum', currscore, step=model.Eiters)
return currscore
def save_checkpoint(state, is_best, filename='checkpoint.pth', prefix=''):
logger = logging.getLogger(__name__)
tries = 2
# deal with unstable I/O. Usually not necessary.
while tries:
try:
# don't save checkpoint
# torch.save(state, prefix + filename)
if is_best:
torch.save(state, os.path.join(prefix, 'model_best.pth'))
except IOError as e:
error = e
tries -= 1
else:
break
logger.info('model save {} failed, remaining {} trials'.format(filename, tries))
if not tries:
raise error
def adjust_learning_rate(opt, optimizer, epoch):
logger = logging.getLogger(__name__)
decay_rate = opt.decay_rate
lr_schedules = opt.lr_schedules
if epoch in lr_schedules:
logger.info('Current epoch num is {}, decrease all lr by 10'.format(epoch, ))
for param_group in optimizer.param_groups:
old_lr = param_group['lr']
new_lr = old_lr * decay_rate
param_group['lr'] = new_lr
logger.info('new lr: {}'.format(new_lr))
def count_params(model):
# The unit is M (million)
model_parameters = filter(lambda p: p.requires_grad, model.parameters())
params = sum([np.prod(p.size()) for p in model_parameters])
params = round(params/(1024**2), 2)
return params
if __name__ == '__main__':
main()