-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathmultichannel_trainer.py
205 lines (160 loc) · 7.59 KB
/
multichannel_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
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
from __future__ import division
import os
import torch
from PIL import Image
from tensorboard_logger import configure, log_value
from torch.autograd import Variable
from torch.utils import data
from torchvision.transforms import Compose, Normalize, ToTensor
from tqdm import tqdm
from argmyparse import get_src_only_training_parser, add_additional_params_to_args, add_img_shape_to_args
from datasets import get_dataset
from loss import CrossEntropyLoss2d
from models.model_util import get_optimizer, fix_batchnorm_when_training, \
get_multichannel_model # check_training
from transform import ReLabel, ToLabel, Scale, RandomSizedCrop, RandomHorizontalFlip, RandomRotation
from util import check_if_done, save_checkpoint, adjust_learning_rate, emphasize_str, get_class_weight_from_file
from util import mkdir_if_not_exist, save_dic_to_json
parser = get_src_only_training_parser()
parser.add_argument('--method', type=str, default="MCD", help="Method Name")
parser.add_argument("--method_detail", type=str, default="MFNet-GateFusion",
choices=["MFNet-AddFusion", "MFNet-ConcatConvFusion", "MFNet-GateFusion"])
parser.add_argument("--inch_list", nargs="+", default=[3, 1], type=int)
args = parser.parse_args()
args = add_additional_params_to_args(args)
args = add_img_shape_to_args(args)
args.savename = "multichannel"
detailed_method = args.method + "-" + args.method_detail
if args.resume:
print("=> loading checkpoint '{}'".format(args.resume))
if not os.path.exists(args.resume):
raise OSError("%s does not exist!" % args.resume)
indir, infn = os.path.split(args.resume)
old_savename = args.savename
args.savename = infn.split("-")[0]
print ("savename is %s (original savename %s was overwritten)" % (args.savename, old_savename))
checkpoint = torch.load(args.resume)
args = checkpoint['args'] # Load args!
[model_g1, model_g2], model_f1 = get_multichannel_model(net_name=args.net, input_ch_list=args.inch_list,
n_class=args.n_class, method=detailed_method, res=args.res,
is_data_parallel=args.is_data_parallel)
optimizer = get_optimizer(list(model_g1.parameters()) + list(model_g2.parameters()) + list(model_f1.parameters()),
opt=args.opt, lr=args.lr, momentum=args.momentum,
weight_decay=args.weight_decay)
model_g1.load_state_dict(checkpoint['g1_state_dict'])
model_g2.load_state_dict(checkpoint['g2_state_dict'])
model_f1.load_state_dict(checkpoint['f1_state_dict'])
optimizer.load_state_dict(checkpoint['optimizer'])
print("=> loaded checkpoint '{}'".format(args.resume))
json_fn = os.path.join(args.outdir, "param_%s_resume.json" % args.savename)
check_if_done(json_fn)
args.machine = os.uname()[1]
save_dic_to_json(args.__dict__, json_fn)
start_epoch = checkpoint['epoch']
else:
[model_g1, model_g2], model_f1 = get_multichannel_model(net_name=args.net, input_ch_list=args.inch_list,
n_class=args.n_class, method=detailed_method, res=args.res,
is_data_parallel=args.is_data_parallel)
optimizer = get_optimizer(list(model_g1.parameters()) + list(model_g2.parameters()) + list(model_f1.parameters()),
opt=args.opt, lr=args.lr, momentum=args.momentum,
weight_decay=args.weight_decay)
args.outdir = os.path.join(args.base_outdir, "%s-%s_only_%sch" % (args.src_dataset, args.split, args.input_ch))
args.pth_dir = os.path.join(args.outdir, "pth")
if args.net in ["fcn", "psp"]:
model_name = "%s-%s-res%s" % (args.savename, args.net, args.res)
else:
model_name = "%s-%s" % (args.savename, args.net)
args.tflog_dir = os.path.join(args.outdir, "tflog", model_name)
mkdir_if_not_exist(args.pth_dir)
mkdir_if_not_exist(args.tflog_dir)
json_fn = os.path.join(args.outdir, "param-%s.json" % model_name)
check_if_done(json_fn)
args.machine = os.uname()[1]
save_dic_to_json(args.__dict__, json_fn)
start_epoch = 0
train_img_shape = tuple([int(x) for x in args.train_img_shape])
img_transform_list = [
Scale(train_img_shape, Image.BILINEAR),
ToTensor(),
# Normalize([.485, .456, .406], [.229, .224, .225])
]
if args.augment:
aug_list = [
RandomRotation(),
# RandomVerticalFlip(), # non-realistic
RandomHorizontalFlip(),
RandomSizedCrop()
]
img_transform_list = aug_list + img_transform_list
img_transform = Compose(img_transform_list)
label_transform = Compose([
Scale(train_img_shape, Image.NEAREST),
ToLabel(),
ReLabel(255, args.n_class - 1),
])
src_dataset = get_dataset(dataset_name=args.src_dataset, split=args.split, img_transform=img_transform,
label_transform=label_transform, test=False, input_ch=args.input_ch)
kwargs = {'num_workers': 1, 'pin_memory': True} if torch.cuda.is_available() else {}
train_loader = torch.utils.data.DataLoader(src_dataset, batch_size=args.batch_size, shuffle=True, **kwargs)
weight = get_class_weight_from_file(n_class=args.n_class, weight_filename=args.loss_weights_file,
add_bg_loss=args.add_bg_loss)
if torch.cuda.is_available():
model_g1.cuda()
model_g2.cuda()
model_f1.cuda()
weight = weight.cuda()
criterion = CrossEntropyLoss2d(weight)
configure(args.tflog_dir, flush_secs=5)
model_g1.train()
model_g2.train()
model_f1.train()
if args.fix_bn:
print (emphasize_str("BN layers are NOT trained!"))
fix_batchnorm_when_training(model_g1)
fix_batchnorm_when_training(model_g2)
fix_batchnorm_when_training(model_f1)
# check_training(model)
for epoch in range(start_epoch, args.epochs):
epoch_loss = 0
for ind, (images, labels) in tqdm(enumerate(train_loader)):
imgs = Variable(images)
lbls = Variable(labels)
if torch.cuda.is_available():
imgs, lbls = imgs.cuda(), lbls.cuda()
# update generator and classifiers by source samples
optimizer.zero_grad()
fet1 = model_g1(imgs[:, :args.inch_list[0], :, :])
fet2 = model_g2(imgs[:, args.inch_list[0]:args.inch_list[0] + args.inch_list[1], :, :])
preds = model_f1(fet1, fet2)
if args.net == "psp":
preds = preds[0]
loss = criterion(preds, lbls)
loss.backward()
c_loss = loss.data[0]
epoch_loss += c_loss
optimizer.step()
if ind % 100 == 0:
print("iter [%d] CLoss: %.4f" % (ind, c_loss))
if ind > args.max_iter:
break
print("Epoch [%d] Loss: %.4f" % (epoch + 1, epoch_loss))
log_value('loss', epoch_loss, epoch)
log_value('lr', args.lr, epoch)
if args.adjust_lr:
args.lr = adjust_learning_rate(optimizer, args.lr, args.weight_decay, epoch, args.epochs)
if args.net == "fcn" or args.net == "psp":
checkpoint_fn = os.path.join(args.pth_dir, "%s-%s-res%s-%s.pth.tar" % (
args.savename, args.net, args.res, epoch + 1))
else:
checkpoint_fn = os.path.join(args.pth_dir, "%s-%s-%s.pth.tar" % (
args.savename, args.net, epoch + 1))
args.start_epoch = epoch + 1
save_dic = {
'args': args,
'epoch': epoch + 1,
'g1_state_dict': model_g1.state_dict(),
'g2_state_dict': model_g2.state_dict(),
'f1_state_dict': model_f1.state_dict(),
'optimizer': optimizer.state_dict()
}
save_checkpoint(save_dic, is_best=False, filename=checkpoint_fn)