-
Notifications
You must be signed in to change notification settings - Fork 1
/
trainer.py
168 lines (138 loc) · 5.79 KB
/
trainer.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
# encoding: utf-8
'''
@author: lwp, syl
'''
import os
import torch
import numpy as np
import utils.utility as utility
from scipy.spatial.distance import cdist
from utils.functions import cmc, mean_ap
from utils.lr_scheduler import WarmupMultiStepLR
from loss.center_loss import CenterLoss
class Trainer():
def __init__(self, args, model, loss, center_criterion, loader, ckpt):
self.args = args
self.train_loader = loader.train_loader
self.test_loader = loader.test_loader
self.query_loader = loader.query_loader
self.testset = loader.testset
self.queryset = loader.queryset
self.ckpt = ckpt
self.model = model
self.loss = loss
self.center_criterion = center_criterion
self.lr = 0.
# center loss
if args.center == 'yes':
self.optimizer, self.optimizer_center = utility.make_optimizer(args, self.model, self.center_criterion)
else:
self.optimizer = utility.make_optimizer(args, self.model, self.center_criterion)
# warmup
if args.warmup == 'yes':
self.scheduler = WarmupMultiStepLR(self.optimizer, args.decay_list, args.gamma, args.warmup_factor, args.warmup_iters, args.warmup_method)
else:
self.scheduler = utility.make_scheduler(args, self.optimizer)
self.device = torch.device('cpu' if args.cpu else 'cuda')
if args.load != '':
self.optimizer.load_state_dict(
torch.load(os.path.join(ckpt.dir, 'optimizer.pt'))
)
for _ in range(len(ckpt.log)*args.test_every): self.scheduler.step()
def train(self):
self.scheduler.step()
self.loss.step()
epoch = self.scheduler.last_epoch + 1
lr = self.scheduler.get_lr()[0]
if lr != self.lr:
self.ckpt.write_log('[INFO] Epoch: {}\tLearning rate: {:.2e}'.format(epoch, lr))
self.lr = lr
self.loss.start_log()
self.model.train()
for batch, (inputs, labels) in enumerate(self.train_loader):
inputs = inputs.to(self.device)
labels = labels.to(self.device)
self.optimizer.zero_grad()
# center loss
if self.args.center == 'yes':
self.optimizer_center.zero_grad()
outputs = self.model(inputs)
loss = self.loss(outputs, labels)
loss.backward()
self.optimizer.step()
# center loss
if self.args.center == 'yes':
for param in self.center_criterion.parameters():
param.grad.data *= (1. / self.args.center_loss_weight)
self.optimizer_center.step()
self.ckpt.write_log('\r[INFO] [{}/{}]\t{}/{}\t{}'.format(
epoch, self.args.epochs,
batch + 1, len(self.train_loader),
self.loss.display_loss(batch)),
end='' if batch+1 != len(self.train_loader) else '\n', progress=True)
self.loss.end_log(len(self.train_loader))
def test(self):
epoch = self.scheduler.last_epoch + 1
self.ckpt.write_log('\n[INFO] Test:')
self.model.eval()
self.ckpt.add_log(torch.zeros(1, 5))
qf_tensor = self.extract_feature(self.query_loader)
gf_tensor = self.extract_feature(self.test_loader)
# no rerank
'''
dist = 1-torch.mm(qf_tensor, gf_tensor.t()) # cosine
dist_np = dist.numpy()
'''
dist_np = cdist(qf_tensor.numpy(), gf_tensor.numpy(), metric='euclidean')
m_ap = mean_ap(dist_np, self.queryset.ids, self.testset.ids, self.queryset.cameras, self.testset.cameras)
r = cmc(dist_np, self.queryset.ids, self.testset.ids, self.queryset.cameras, self.testset.cameras,
separate_camera_set=False,
single_gallery_shot=False,
first_match_break=True)
self.ckpt.log[-1, 0] = m_ap
self.ckpt.log[-1, 1] = r[0]
self.ckpt.log[-1, 2] = r[2]
self.ckpt.log[-1, 3] = r[4]
self.ckpt.log[-1, 4] = r[9]
best = self.ckpt.log.max(0)
self.ckpt.write_log(
'[INFO]( ^_^ ) mAP: {:.4f} rank1: {:.4f} rank3: {:.4f} rank5: {:.4f} rank10: {:.4f} (Best: {:.4f} @epoch {})'.format(
m_ap,
r[0], r[2], r[4], r[9],
best[0][0],
(best[1][0] + 1)*self.args.test_every
)
)
if not self.args.test_only:
self.ckpt.save(self, epoch, is_best=((best[1][0] + 1)*self.args.test_every == epoch))
def fliphor(self, inputs):
'''
flip horizontal
'''
inv_idx = torch.arange(inputs.size(3)-1,-1,-1).long() # N x C x H x W
img_flip = inputs.index_select(3,inv_idx)
return img_flip
def L2Normalization(self, ff, dim): # add by lwp, 2020.8.20
fnorm = torch.norm(ff, p=2, dim=1, keepdim=True)
ff = ff.div(fnorm.expand_as(ff))
return ff
def extract_feature(self, loader): # modified by lwp, 2020.8.20
features = torch.FloatTensor()
for (inputs, labels) in loader:
ff = torch.FloatTensor(inputs.size(0), 4096).zero_().cuda()
for i in range(2):
if i==1:
inputs = self.fliphor(inputs)
input_img = inputs.to(self.device)
outputs = self.model(input_img)
ff += outputs[0]
ff = self.L2Normalization(ff, dim=1)
features = torch.cat((features, ff.data.cpu().float()), 0)
return features
def terminate(self):
if self.args.test_only:
self.test()
return True
else:
epoch = self.scheduler.last_epoch + 1
return epoch >= self.args.epochs