-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathtrain.py
241 lines (202 loc) · 10.3 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
#coding=utf-8
import argparse
import os
import time
import logging
import random
import numpy as np
from collections import OrderedDict
import setproctitle
import torch
import torch.optim
import torch.nn.functional as F
import torch.backends.cudnn as cudnn
from torch.utils.data import DataLoader
from torch.utils.tensorboard import SummaryWriter
from data.transforms import *
# from data.datasets_nii import Brats_loadall_nii, Brats_loadall_test_nii
from data.BraTS import BraTS
from data.data_utils import init_fn
from modules.fusionseg import Model
from utils import Parser, criterions
from utils.parser import setup
from utils.lr_scheduler import LR_Scheduler, record_loss, MultiEpochsDataLoader
# from predict import AverageMeter, test_softmax
local_time = time.strftime("%Y%m%d %H%M%S", time.localtime())
parser = argparse.ArgumentParser()
parser.add_argument('--user', default='lizy', type=str)
# setting
parser.add_argument('-batch_size', '--batch_size', default=2, type=int, help='Batch size')
parser.add_argument('--project_root', default='/home/l/data_2/lzy/Bra18/EATE-main', type=str)
# parser.add_argument('--datapath', default='/home/a/data/lzy/Dataset/BRATS2018_Training_none_npy', type=str)
parser.add_argument('--root', default='/home/l/data_1/lzy/BraTSDataset/BraTS2018/MICCAI_BraTS_2018_Data_Training', type=str)
parser.add_argument('--train_file', default='/home/l/data_1/lzy/BraTSDataset/BraTS2018/train.txt', type=str)
parser.add_argument('--dataname', default='BRATS2018', type=str)
parser.add_argument('--savepath', default='/home/l/data_2/lzy/Bra18/EATE-main/output', type=str)
parser.add_argument('--resume', default='/home/l/data_2/lzy/Bra18/EATE-main/output/model_1130.pth', type=str)
parser.add_argument('--num_epochs', default=2000, type=int)
parser.add_argument('--start_epoch', default=0, type=int)
parser.add_argument('--lr', default=2e-4, type=float)
parser.add_argument('--gpu', default='0,1', type=str)
parser.add_argument('--load', default=False, type=bool) # 加载模型
parser.add_argument('--num_cls', default=4, type=int)
parser.add_argument('--date', default=local_time.split(' ')[0], type=str)
parser.add_argument('--pretrain', default=None, type=str)
parser.add_argument('--weight_decay', default=1e-4, type=float)
parser.add_argument('--iter_per_epoch', default=150, type=int)
parser.add_argument('--region_fusion_start_epoch', default=0, type=int)
parser.add_argument('--seed', default=1024, type=int)
path = os.path.dirname(__file__)
## parse arguments
args = parser.parse_args()
setup(args, 'training')
args.train_transforms = 'Compose([RandCrop3D((128,128,128)), RandomRotion(10), RandomIntensityChange((0.1,0.1)), RandomFlip(0), NumpyType((np.float32, np.int64)),])'
args.test_transforms = 'Compose([NumpyType((np.float32, np.int64)),])'
ckpts = args.savepath
os.makedirs(ckpts, exist_ok=True)
###tensorboard writer
writer = SummaryWriter(os.path.join(args.savepath, 'summary'))
def main():
##########setting gpu
os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu
##########setting seed
torch.manual_seed(args.seed)
torch.cuda.manual_seed(args.seed)
torch.cuda.manual_seed_all(args.seed)
random.seed(args.seed)
np.random.seed(args.seed)
cudnn.benchmark = False
cudnn.deterministic = True
##########setting models
model = Model(num_cls=args.num_cls)
# print (model)
model = torch.nn.DataParallel(model).cuda()
##########Setting learning schedule and optimizer
lr_schedule = LR_Scheduler(args.lr, args.num_epochs)
train_params = [{'params': model.parameters(), 'lr': args.lr, 'weight_decay': args.weight_decay}]
optimizer = torch.optim.Adam(train_params, betas=(0.9, 0.999), eps=1e-08, amsgrad=True)
logging.info(str(args))
if os.path.isfile(args.resume) and args.load:
logging.info('loading checkpoint {}'.format(args.resume))
checkpoint = torch.load(args.resume)
model.load_state_dict(checkpoint['state_dict'])
logging.info('successfully loading checkpoint {} and traing from epoch:{}'.format(args.resume, args.start_epoch))
else:
logging.info('re-traing!')
# train_set = Brats_loadall_nii(transforms=args.train_transforms, root=args.datapath, num_cls=args.num_cls)
train_set = BraTS(args.train_file, args.root, mode='train')
train_loader = MultiEpochsDataLoader(
dataset=train_set,
batch_size=args.batch_size,
num_workers=8,
pin_memory=True,
shuffle=True,
worker_init_fn=init_fn)
##########Training
start = time.time()
torch.set_grad_enabled(True)
logging.info('#############training############')
# iter_per_epoch = args.iter_per_epoch
iter_per_epoch = len(train_loader)
train_iter = iter(train_loader)
for epoch in range(args.start_epoch, args.num_epochs):
setproctitle.setproctitle('{}: {}/{}'.format(args.user, epoch + 1, args.num_epochs))
step_lr = lr_schedule(optimizer, epoch)
writer.add_scalar('lr', step_lr, global_step=(epoch+1))
b = time.time()
for i in range(iter_per_epoch):
step = (i+1) + epoch*iter_per_epoch
###Data load
try:
data = next(train_iter)
except:
train_iter = iter(train_loader)
data = next(train_iter)
x, target = data[:2] # data的前2个数据
x = x.cuda(non_blocking=True)
target = target.cuda(non_blocking=True)
# model.module.is_training = True
final_out, out1, outs2, outs3, masks1, masks2 = model(x)
###Loss compute
"""final loss"""
final_cross_loss = criterions.softmax_weighted_loss(final_out, target, num_cls=args.num_cls)
final_dice_loss = criterions.dice_loss(final_out, target, num_cls=args.num_cls)
final_loss = final_cross_loss + final_dice_loss
'''out1'''
out1_cross_loss = criterions.softmax_weighted_loss(out1, target, num_cls=args.num_cls)
out1_dice_loss = criterions.dice_loss(out1, target, num_cls=args.num_cls)
out1_loss = out1_cross_loss + out1_dice_loss
'''outs2'''
"""masks loss"""
out2_cross_loss = torch.zeros(1).cuda().float()
out2_dice_loss = torch.zeros(1).cuda().float()
for out2 in outs2:
out2_cross_loss += criterions.softmax_weighted_loss(out2, target, num_cls=args.num_cls)
out2_dice_loss += criterions.dice_loss(out2, target, num_cls=args.num_cls)
out2_loss = out2_cross_loss + out2_dice_loss
"""outs3 loss"""
out3_cross_loss = torch.zeros(1).cuda().float()
out3_dice_loss = torch.zeros(1).cuda().float()
for out3 in outs3:
out3_cross_loss += criterions.softmax_weighted_loss(out3, target, num_cls=args.num_cls)
out3_dice_loss += criterions.dice_loss(out3, target, num_cls=args.num_cls)
out3_loss = out3_cross_loss + out3_dice_loss
"""mask1"""
mask1_cross_loss = torch.zeros(1).cuda().float()
mask1_dice_loss = torch.zeros(1).cuda().float()
for mask1 in masks1:
mask1_cross_loss += criterions.softmax_weighted_loss(mask1, target, num_cls=args.num_cls)
mask1_dice_loss += criterions.dice_loss(mask1, target, num_cls=args.num_cls)
mask1_loss = mask1_cross_loss + mask1_dice_loss
"""fusion_preds loss"""
mask2_cross_loss = torch.zeros(1).cuda().float()
mask2_dice_loss = torch.zeros(1).cuda().float()
for mask2 in masks2:
mask2_cross_loss += criterions.softmax_weighted_loss(mask2, target, num_cls=args.num_cls)
mask2_dice_loss += criterions.dice_loss(mask2, target, num_cls=args.num_cls)
mask2_loss = mask2_cross_loss + mask2_dice_loss
"""total loss"""
loss = final_loss + out1_loss + out2_loss + out3_loss + mask1_loss + mask2_loss
optimizer.zero_grad()
loss.backward()
optimizer.step()
###log
writer.add_scalar('loss', loss.item(), global_step=step)
writer.add_scalar('final_cross_loss', final_cross_loss.item(), global_step=step)
writer.add_scalar('final_dice_loss', final_dice_loss.item(), global_step=step)
writer.add_scalar('out1_dice_loss', out1_dice_loss.item(), global_step=step)
writer.add_scalar('out2_dice_loss', out2_dice_loss.item(), global_step=step)
writer.add_scalar('out3_dice_loss', out3_dice_loss.item(), global_step=step)
writer.add_scalar('mask1_dice_loss', mask1_dice_loss.item(), global_step=step)
writer.add_scalar('mask2_dice_loss', mask2_dice_loss.item(), global_step=step)
msg = 'Epoch {}/{}, Iter {}/{}, Loss {:.4f}, '.format((epoch + 1), args.num_epochs, (i + 1), iter_per_epoch,
loss.item())
msg += 'finaldice:{:.4f}, cross:{:.4f},'.format(final_dice_loss.item(), final_cross_loss.item())
msg += 'out1:{:.4f},:{:.4f},'.format(out1_dice_loss.item(), out1_cross_loss.item())
msg += 'out2:{:.4f},:{:.4f},'.format(out2_dice_loss.item(), out2_cross_loss.item())
msg += 'out3:{:.4f},:{:.4f},'.format(out3_dice_loss.item(), out3_cross_loss.item())
msg += 'mask1:{:.4f},:{:.4f},'.format(mask1_dice_loss.item(), mask1_cross_loss.item())
msg += 'mask2:{:.4f},:{:.4f}'.format(mask2_dice_loss.item(), mask2_cross_loss.item())
logging.info(msg)
logging.info('train time per epoch: {}'.format(time.time() - b))
##########model save
file_name = os.path.join(ckpts, 'model_last.pth')
torch.save({
'epoch': epoch,
'state_dict': model.state_dict(),
'optim_dict': optimizer.state_dict(),
},
file_name)
if (epoch+1) % 10 == 0 or (epoch>=(args.num_epochs-10)):
file_name = os.path.join(ckpts, 'model_{}.pth'.format(epoch+1))
torch.save({
'epoch': epoch,
'state_dict': model.state_dict(),
'optim_dict': optimizer.state_dict(),
},
file_name)
msg = 'total time: {:.4f} hours'.format((time.time() - start)/3600)
logging.info(msg)
if __name__ == '__main__':
main()
# python train.py