-
Notifications
You must be signed in to change notification settings - Fork 6
/
train.py
356 lines (306 loc) · 17 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
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
import os
os.environ["KMP_DUPLICATE_LIB_OK"]="TRUE"
import glob
import logging
import random
from dataset import *
from losses import *
from models import SEANet
from tensorboardX import SummaryWriter
from torch import nn
from torch.utils.data import DataLoader
from tqdm import tqdm
from utils import visualize, evaluate, create_train_arg_parser
import pickle
import numpy as np
from collections import defaultdict
from torch.optim import lr_scheduler
import scipy.io as sio
from torchtoolbox.tools import mixup_data, mixup_criterion
from sklearn.model_selection import train_test_split
# from albumentations.augmentations import transforms
# from albumentations.core.composition import Compose, OneOf
# from albumentations import RandomRotate90,Resize,Cutout
# from thop import profile
# from torchinfo import summary
##权重初始化
def weights_init(m):
if isinstance(m, nn.Conv2d):
# xavier(m.weight.data)
m.weight.data.normal_(0, 0.01)
if m.weight.data.shape == torch.Size([1, 5, 1, 1]):
# for new_score_weight
torch.nn.init.constant_(m.weight, 0.2)
if m.bias is not None:
m.bias.data.zero_()
#RCF预训练文件
def load_vgg16pretrain(model, vggmodel='./vgg16convs.mat'):
vgg16 = sio.loadmat(vggmodel)
torch_params = model.state_dict()
for k in vgg16.keys():
name_par = k.split('-')
size = len(name_par)
if size == 2:
name_space = name_par[0] + '.' + name_par[1]
data = np.squeeze(vgg16[k])
torch_params[name_space] = torch.from_numpy(data)
model.load_state_dict(torch_params)
def build_model(model_type):
if model_type == "rcf":
model = SEANet(num_classes=1)
return model
if __name__ == "__main__":
args = create_train_arg_parser().parse_args()
args.distance_type = 'dist_contour'
args.train_path = r'D:\LJ2\SBA2\DM_5706\image'
# args.val_path = './XJ_kerl_2/valid/image/'
args.model_type = 'rcf'
args.object_type = 'polyp'
args.save_path = r'D:\LJ2\SBA2\DM_5706\se_dm'
# args.pretrained_model_path = './XJ_new_model_xinzeng/15.pt'
CUDA_SELECT = "cuda:{}".format(args.cuda_no)
log_path = args.save_path + "/summary"
writer = SummaryWriter(log_dir=log_path)
logging.basicConfig(
filename="".format(args.object_type),
filemode="a",
format="%(asctime)s,%(msecs)d %(name)s %(levelname)s %(message)s",
datefmt="%Y-%m-%d %H:%M",
level=logging.INFO,
)
logging.info("")
train_file_names = glob.glob(os.path.join(args.train_path, "*.tif"))
random.shuffle(train_file_names)
img_ids = [os.path.splitext(os.path.basename(p))[0] for p in train_file_names]
train_file, val_file = train_test_split(img_ids, test_size=0.2, random_state=41)
device = torch.device(CUDA_SELECT if torch.cuda.is_available() else "cpu")
print(device)
model = build_model(args.model_type)
# print(summary(model,(1,3,256,256)))
if torch.cuda.device_count() > 0: #本来是0
print("Let's use", torch.cuda.device_count(), "GPUs!")
#model = nn.DataParallel(model)
model = model.to(device)
# To handle epoch start number and pretrained weight
# if args.vgg16_caffe:
# load_vgg16_caffe(model, args.vgg16_caffe) HED
model.apply(weights_init)
load_vgg16pretrain(model) #
# net_parameters_id = {}
net_parameters_id = defaultdict(list)
net = model
for pname, p in net.named_parameters():
print(pname)
if pname in ['conv1_1.weight','conv1_2.weight',
'conv2_1.weight','conv2_2.weight',
'conv3_1.weight','conv3_2.weight','conv3_3.weight',
'conv4_1.weight','conv4_2.weight','conv4_3.weight']:
# print(pname, 'lr:1 de:1')
if 'conv1-4.weight' not in net_parameters_id:
net_parameters_id['conv1-4.weight'] = []
net_parameters_id['conv1-4.weight'].append(p)
elif pname in ['conv_final1.weight','conv_final2.weight']:
# print(pname, 'lr:2 de:0')
net_parameters_id['final1-2.weight'].append(p)
elif pname in ['conv_final1.bias','conv_final2.bias']:
# print(pname, 'lr:2 de:0')
net_parameters_id['final1-2.bias'].append(p)
elif pname in ['conv1_1.bias','conv1_2.bias',
'conv2_1.bias','conv2_2.bias',
'conv3_1.bias','conv3_2.bias','conv3_3.bias',
'conv4_1.bias','conv4_2.bias','conv4_3.bias']:
# print(pname, 'lr:2 de:0')
if 'conv1-4.bias' not in net_parameters_id:
net_parameters_id['conv1-4.bias'] = []
net_parameters_id['conv1-4.bias'].append(p)
elif pname in ['conv5_1.weight','conv5_2.weight','conv5_3.weight']:
# print(pname, 'lr:100 de:1')
if 'conv5.weight' not in net_parameters_id:
net_parameters_id['conv5.weight'] = []
net_parameters_id['conv5.weight'].append(p)
elif pname in ['conv5_1.bias','conv5_2.bias','conv5_3.bias'] :
# print(pname, 'lr:200 de:0')
if 'conv5.bias' not in net_parameters_id:
net_parameters_id['conv5.bias'] = []
net_parameters_id['conv5.bias'].append(p)
elif pname in ['conv1_1_down.weight','conv1_2_down.weight',
'conv2_1_down.weight','conv2_2_down.weight',
'conv3_1_down.weight','conv3_2_down.weight','conv3_3_down.weight',
'conv4_1_down.weight','conv4_2_down.weight','conv4_3_down.weight',
'conv5_1_down.weight','conv5_2_down.weight','conv5_3_down.weight']:
# print(pname, 'lr:0.1 de:1')
if 'conv_down_1-5.weight' not in net_parameters_id:
net_parameters_id['conv_down_1-5.weight'] = []
net_parameters_id['conv_down_1-5.weight'].append(p)
elif pname in ['conv1_1_down.bias','conv1_2_down.bias',
'conv2_1_down.bias','conv2_2_down.bias',
'conv3_1_down.bias','conv3_2_down.bias','conv3_3_down.bias',
'conv4_1_down.bias','conv4_2_down.bias','conv4_3_down.bias',
'conv5_1_down.bias','conv5_2_down.bias','conv5_3_down.bias']:
# print(pname, 'lr:0.2 de:0')
if 'conv_down_1-5.bias' not in net_parameters_id:
net_parameters_id['conv_down_1-5.bias'] = []
net_parameters_id['conv_down_1-5.bias'].append(p)
elif pname in ['score_dsn1.weight','score_dsn2.weight','score_dsn3.weight',
'score_dsn4.weight','score_dsn5.weight']:
# print(pname, 'lr:0.01 de:1')
if 'score_dsn_1-5.weight' not in net_parameters_id:
net_parameters_id['score_dsn_1-5.weight'] = []
net_parameters_id['score_dsn_1-5.weight'].append(p)
elif pname in ['score_dsn1.bias','score_dsn2.bias','score_dsn3.bias',
'score_dsn4.bias','score_dsn5.bias']:
# print(pname, 'lr:0.02 de:0')
if 'score_dsn_1-5.bias' not in net_parameters_id:
net_parameters_id['score_dsn_1-5.bias'] = []
net_parameters_id['score_dsn_1-5.bias'].append(p)
elif pname in ['score_final.weight']:
# print(pname, 'lr:0.001 de:1')
if 'score_final.weight' not in net_parameters_id:
net_parameters_id['score_final.weight'] = []
net_parameters_id['score_final.weight'].append(p)
elif pname in ['score_final.bias']:
# print(pname, 'lr:0.002 de:0')
if 'score_final.bias' not in net_parameters_id:
net_parameters_id['score_final.bias'] = []
net_parameters_id['score_final.bias'].append(p)
elif pname in ['aspp.convs.0.0.weight','aspp.convs.0.1.weight','aspp.convs.1.0.weight','aspp.convs.1.1.weight','aspp.convs.2.0.weight',
'aspp.convs.2.1.weight','aspp.convs.3.0.weight','aspp.convs.3.1.weight','aspp.convs.4.1.weight','aspp.convs.4.2.weight',
'aspp.project.0.weight','aspp.project.1.weight']:
# print(pname, 'lr:0.002 de:0')
if 'aspp1-12.weight' not in net_parameters_id:
net_parameters_id['aspp1-12.weight'] = []
net_parameters_id['aspp1-12.weight'].append(p)
elif pname in ['aspp.convs.0.1.bias','aspp.convs.1.1.bias','aspp.convs.2.1.bias','aspp.convs.3.1.bias',
'aspp.convs.4.2.bias','aspp.project.1.bias']:
# print(pname, 'lr:0.002 de:0')
if 'aspp1-6.bias' not in net_parameters_id:
net_parameters_id['aspp1-6.bias'] = []
net_parameters_id['aspp1-6.bias'].append(p)
elif pname in ['aspp1.convs.0.0.weight','aspp1.convs.0.1.weight','aspp1.convs.1.0.weight','aspp1.convs.1.1.weight','aspp1.convs.2.0.weight',
'aspp1.convs.2.1.weight','aspp1.convs.3.0.weight','aspp1.convs.3.1.weight','aspp1.convs.4.1.weight','aspp1.convs.4.2.weight',
'aspp1.project.0.weight','aspp1.project.1.weight']:
# print(pname, 'lr:0.002 de:0')
if 'as1-12.weight' not in net_parameters_id:
net_parameters_id['as1-12.weight'] = []
net_parameters_id['as1-12.weight'].append(p)
elif pname in ['aspp1.convs.0.1.bias','aspp1.convs.1.1.bias','aspp1.convs.2.1.bias','aspp1.convs.3.1.bias',
'aspp1.convs.4.2.bias','aspp1.project.1.bias']:
# print(pname, 'lr:0.002 de:0')
if 'as1-6.bias' not in net_parameters_id:
net_parameters_id['as1-6.bias'] = []
net_parameters_id['as1-6.bias'].append(p)
elif pname in ['center.block.1.conv.weight','dec5.block.1.conv.weight','dec5.block.2.conv.weight','dec4.block.1.conv.weight','dec4.block.2.conv.weight','dec3.block.2.conv.weight','dec3.block.1.conv.weight',
'dec2.block.2.conv.weight','dec2.block.1.conv.weight','dec1.conv.weight',' xdf.conv.weight','center.block.2.conv.weight',
'center.block.3.fc.0.weight','center.block.3.fc.2.weight','dec5.block.3.fc.0.weight','dec5.block.3.fc.2.weight','dec4.block.3.fc.0.weight',
'dec4.block.3.fc.2.weight','dec3.block.3.fc.0.weight','dec3.block.3.fc.2.weight','dec2.block.3.fc.0.weight','dec2.block.3.fc.2.weight']:
# print(pname, 'lr:0.002 de:0')
if 'dec1-22.weight' not in net_parameters_id:
net_parameters_id['dec1-22.weight'] = []
net_parameters_id['dec1-22.weight'].append(p)
elif pname in ['center.block.1.conv.bias','dec5.block.1.conv.bias','dec5.block.2.conv.bias','dec4.block.1.conv.bias','dec4.block.2.conv.bias','dec3.block.2.conv.bias','dec3.block.1.conv.bias',
'dec2.block.2.conv.bias','dec2.block.1.conv.bias','dec1.conv.bias',' xdf.conv.bias','center.block.2.conv.bias']:
# print(pname, 'lr:0.002 de:0')
if 'dec1-12.bias' not in net_parameters_id:
net_parameters_id['dec1-12.bias'] = []
net_parameters_id['dec1-12.bias'].append(p)
optimizer = torch.optim.SGD([
{'params': net_parameters_id['conv1-4.weight'] , 'lr': args.lr*1 , 'weight_decay': args.weight_decay},
{'params': net_parameters_id['conv1-4.bias'] , 'lr': args.lr*2 , 'weight_decay': 0.},
{'params': net_parameters_id['conv5.weight'] , 'lr': args.lr*10 , 'weight_decay': args.weight_decay},
{'params': net_parameters_id['conv5.bias'] , 'lr': args.lr*20 , 'weight_decay': 0.},
{'params': net_parameters_id['conv_down_1-5.weight'], 'lr': args.lr*0.1 , 'weight_decay': args.weight_decay},
{'params': net_parameters_id['conv_down_1-5.bias'] , 'lr': args.lr*0.2 , 'weight_decay': 0.},
{'params': net_parameters_id['score_dsn_1-5.weight'], 'lr': args.lr*0.01 , 'weight_decay': args.weight_decay},
{'params': net_parameters_id['score_dsn_1-5.bias'] , 'lr': args.lr*0.02 , 'weight_decay': 0.},
{'params': net_parameters_id['score_final.weight'] , 'lr': args.lr*0.001, 'weight_decay': args.weight_decay},
{'params': net_parameters_id['score_final.bias'] , 'lr': args.lr*0.002, 'weight_decay': 0.},
{'params': awl.parameters(),'lr': args.lr}
], lr=args.lr, momentum=args.momentum, weight_decay=args.weight_decay)
scheduler = lr_scheduler.StepLR(optimizer, step_size=args.stepsize, gamma=args.gamma)
epoch_start = "0"
if args.use_pretrained:
print("Loading Model {}".format(os.path.basename(args.pretrained_model_path)))
model.load_state_dict(torch.load(args.pretrained_model_path))
epoch_start = os.path.basename(args.pretrained_model_path).split(".")[0]
print(epoch_start)
#print(args.use_pretrained)
trainLoader = DataLoader(
DatasetImageMaskContourDist(args.train_path,train_file, args.distance_type),
batch_size=args.batch_size,drop_last=False, shuffle=True
)
devLoader = DataLoader(
DatasetImageMaskContourDist(args.train_path,val_file, args.distance_type),drop_last=False,shuffle=False
)
displayLoader = DataLoader(
DatasetImageMaskContourDist(args.train_path,val_file, args.distance_type),
batch_size=args.val_batch_size,drop_last=False,shuffle=False
)
optimizer1 = Apollo([{'params': net_parameters_id['final1-2.weight'],'lr': args.lr1},
{'params': net_parameters_id['dec1-22.weight'],'lr': args.lr1},
{'params': net_parameters_id['dec1-12.bias'], 'lr': args.lr1},
{'params': net_parameters_id['final1-2.bias'], 'lr': args.lr1},
{'params': net_parameters_id['aspp1-12.weight'], 'lr': args.lr1},
{'params': net_parameters_id['aspp1-6.bias'], 'lr': args.lr1},
{'params': net_parameters_id['as1-12.weight'], 'lr': args.lr1},
{'params': net_parameters_id['as1-6.bias'], 'lr': args.lr1}], lr=args.lr1)
# scheduler1 = lr_scheduler.StepLR(optimizer1, step_size=args.stepsize, gamma=args.gamma)
# scheduler1 = torch.optim.lr_scheduler.CosineAnnealingLR(optimizer1, int(1e10), eta_min=1e-5)
for epoch in tqdm(
range(int(epoch_start) + 1, int(epoch_start) + 1 + args.num_epochs)
):
global_step = epoch * len(trainLoader)
running_loss = 0.0
counter = 0
for i, (img_file_name, inputs, targets1, targets2,targets3) in enumerate(
tqdm(trainLoader)
):
model.train()
inputs = inputs.to(device)
targets1 = targets1.to(device).float()
targets2 = targets2.to(device).float()
targets3 = targets3.to(device).float()
counter += 1
###mix_up，from torchtoolbox.tools import mixup_data, mixup_criterion（pip install torchtoolbox)
# alpha=0.2
# data1, labels_a, labels_b, lam1 = mixup_data(inputs, targets1, alpha)
# data2, labels_c, labels_d, lam2 = mixup_data(inputs, targets2, alpha)
# data3, labels_e, labels_f, lam3 = mixup_data(inputs, targets3, alpha)
criterion3 = nn.MSELoss()
criterion1 = BCEDiceLoss()
criterion2 = RCFloss()
with torch.set_grad_enabled(True):
output1, output2, output3 = model(inputs.float())
loss1 = criterion1(output1,targets1)
loss2 = criterion2(output2,targets2)
loss3 = criterion3(output3,targets3)
# loss1 = mixup_criterion(criterion1, output1, labels_a, labels_b, lam1)
# loss2 = mixup_criterion(criterion2, output2, labels_c, labels_d, lam2)
# loss3 = mixup_criterion(criterion3, output3, labels_e, labels_f, lam3)
loss = awl(loss1,loss3)
loss = loss + loss2
loss = loss / args.itersize
loss.backward()
if counter == args.itersize:
optimizer.step()
optimizer.zero_grad()
optimizer1.step()
optimizer1.zero_grad()
counter = 0
writer.add_scalar("loss", loss.item(), epoch)
running_loss += loss.item() * inputs.size(0)
epoch_loss = running_loss / len(train_file_names)
scheduler.step()
# scheduler1.step()
# print(epoch_loss)
if epoch % 1 == 0:
dev_loss, dev_time = evaluate(device, epoch, model, devLoader, writer)
writer.add_scalar("valid overall accuracy", dev_loss, epoch)
visualize(device, epoch, model, displayLoader, writer, args.val_batch_size)
print("Global Loss:{} Val Loss:{}".format(epoch_loss, dev_loss))
else:
print("Global Loss:{} ".format(epoch_loss))
logging.info("epoch:{} train_loss:{} ".format(epoch, epoch_loss))
if epoch % 5 == 0:
torch.save(
model.state_dict(), os.path.join(args.save_path, str(epoch) + ".pt")
)