-
Notifications
You must be signed in to change notification settings - Fork 0
/
main.py
219 lines (177 loc) · 7.85 KB
/
main.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
import torch
import torch.nn as nn
import torch.optim as optim
import random
import torch.backends.cudnn as cudnn
from torch.utils.data import DataLoader
import time
import os
from torch.autograd import Variable
from PIL import Image
import cv2
from model import *
from config import *
from dataset import FlameSet,DataPrefetcher
def main():
print(opt)
cuda = opt.cuda
if cuda and not torch.cuda.is_available():
raise Exception("No GPU found, please run without --cuda")
opt.seed = random.randint(1, 10000)
torch.manual_seed(opt.seed)
if cuda:
torch.cuda.manual_seed(opt.seed)
cudnn.benchmark = True
print('======>Load datasets')
train_set = FlameSet(opt.dataset_train)
val_set = FlameSet(opt.dataset_val)
train_data_loader = DataLoader(dataset=train_set,batch_size=opt.batchSize,shuffle=True)
val_data_loader = DataLoader(dataset=val_set, batch_size=opt.batchSize, shuffle=True)
print("===> Building model")
model = MPFINet()
criterion = nn.L1Loss()
min_loss = 100.
if cuda:
model = model.cuda()
criterion = criterion.cuda()
print("===> Setting GPU Cuda")
print("===> Setting Optimizer and lr_scheduler")
optimizer = optim.AdamW(model.parameters(), lr=opt.lr,betas=[0.9,0.99])
scheduler = torch.optim.lr_scheduler.MultiStepLR(optimizer,milestones=[301,601],gamma=0.1)
print("===> Parameter numbers : %.2fM" % (sum(p.numel() for p in model.parameters()) / 1000000.0))
# optionally resume from a checkpoint
if opt.resume:
if os.path.isfile(opt.resume):
print("===> loading checkpoint '{}'".format(opt.resume))
checkpoint = torch.load(opt.resume)
opt.start_epoch = checkpoint["epoch"] + 1
model.load_state_dict(checkpoint["model_state_dict"])
optimizer.load_state_dict((checkpoint["optimizer_state_dict"]))
else:
print("===> no checkpoint found at '{}'".format(opt.resume))
# optionally copy weights from a checkpoint
if opt.pretrained:
if os.path.isfile(opt.pretrained):
print("===> loading model '{}'".format(opt.pretrained))
weights = torch.load(opt.pretrained)
model.load_state_dict(weights['model'].state_dict())
else:
print("===> no model found at '{}'".format(opt.pretrained))
print("===> Training")
t = time.strftime("%Y%m%d%H%M")
if not os.path.exists('log'):
os.mkdir('log')
global log
log = 'log/'+t+ '_batchSize%d' % (opt.batchSize)+'_lr%f'% (opt.lr)+'_QB'
if not os.path.exists(log):
os.mkdir(log)
backup_model = os.path.join(log,opt.backup)
if not os.path.exists(backup_model):
os.mkdir(backup_model)#备份pkl文件
best_model = os.path.join(log,opt.best_model)
if not os.path.exists(best_model):
os.mkdir(best_model)#存储最好结果pkl
train(train_data_loader,val_data_loader, optimizer, model, scheduler,criterion,t,backup_model,best_model,min_loss)
# def adjust_learning_rate(optimizer, epoch):
# """Sets the learning rate to the initial LR decayed by 10"""
# lr = opt.lr * (0.1 ** (epoch // 300))
# return lr
def train(train_data_loader,val_data_loader, optimizer, model, scheduler,criterion,t,backup_model,best_model,min_loss):
print('===> Begin Training!')
model.train()
steps_per_epoch = len(train_data_loader)
if opt.train_log:
if os.path.isfile(opt.train_log):
train_log = open(opt.train_log,"a") # log/Networkname_202111022144_train.log
else:
print("=> no train_log found at '{}'".format(opt.train_log))
else:
train_log = open(os.path.join(log, "%s_%s_train.log") % (opt.net, t),"w") # log/Networkname_202111022144_train.log
if opt.epoch_time_log:
if os.path.isfile(opt.epoch_time_log):
epoch_time_log = open(opt.epoch_time_log,"a")
else:
print("==> no epoch_time_log found at '{}'".format(opt.epoch_time_log))
else:
epoch_time_log = open(os.path.join(log, "%s_%s_epoch_time.log") % (opt.net, t), "w")
time_sum = 0
global val_log
if opt.val_log:
if os.path.isfile(opt.val_log):
val_log = open(opt.val_log, "a")
else:
print("===> no val_log found at '{}'".format(opt.val_log))
else:
val_log = open(os.path.join(log, "%s_%s_val.log") % (opt.net, t), "w")
for epoch in range(opt.start_epoch, opt.nEpochs+1):
# 计时,每个epoch的时间
start = time.time()
#加快读取数据集
prefetcher = DataPrefetcher(train_data_loader)
data = prefetcher.next()
i = 0
while data is not None:
# run_step()
i += 1
if i >= steps_per_epoch:
break # 循环次数从0开始计数还是从1开始计数的问题
input_pan, input_ms, target = data[0], data[1],data[2]
if opt.cuda:
input_pan = input_pan.cuda()
input_ms = input_ms.cuda()
target = target.cuda()
output = model(input_pan, input_ms)
train_loss = criterion(output, target)
optimizer.zero_grad()
train_loss.backward()
optimizer.step()
scheduler.step()#调整lr
print("===> Epoch[{}/{}]: Lr:{} Train_Loss: {:.10f} ".format(epoch,opt.nEpochs,optimizer.param_groups[0]["lr"],train_loss.item()))
train_log.write("Epoch[{}/{}]: Train_Loss: {:.15f}\n".format(epoch, opt.nEpochs, train_loss.item()))
if epoch % 50 ==0:
save_checkpoint(model, epoch, optimizer, backup_model)
# backup a model every ** epochs and validate
if epoch % 2 == 0:
val_retloss = val(val_data_loader, model, criterion, epoch)
val_log.write("{} {:.10f}\n".format((epoch), val_retloss))
# Save the best weight
if min_loss > val_retloss:
save_bestmodel(model, epoch, best_model)
min_loss = val_retloss
# 输出每轮训练花费时间
time_epoch = (time.time() - start)
time_sum += time_epoch
epoch_time_log.write("No:{} epoch training costs {:.4f}min\n".format(epoch, time_epoch / 60))
train_log.close()
val_log.close()
epoch_time_log.close()
def val(val_data_loader, model, criterion, epoch):
model.eval()
avg_l1 = 0
val_loss=[]
with torch.no_grad():
for k, data in enumerate(val_data_loader):
input_pan, input_ms, target = data[0], data[1],data[2]
if opt.cuda:
input_pan = input_pan.cuda()
input_ms = input_ms.cuda()
target = target.cuda()
output = model(input_pan, input_ms)
loss = criterion(output, target)
avg_l1 += loss.item()
del (input_pan, input_ms,target, output)
print("===> Epoch{} Avg. Val Loss: {:.10f}".format(epoch, avg_l1 / len(val_data_loader)))#ssim_value
val_loss.append(avg_l1)
return avg_l1 / len(val_data_loader)
def save_checkpoint(model, epoch,optimizer,backup_model):
model_out_path = os.path.join(backup_model,"model_epoch_last.pth".format(epoch))
state = {"epoch": epoch, "model_state_dict": model.state_dict(),"optimizer_state_dict":optimizer.state_dict()}
torch.save(state, model_out_path)
print("===> Checkpoint saved to {}".format(epoch))
def save_bestmodel(model, epoch,backupbest_model):
model_out_path = os.path.join(backupbest_model,"model_best_epoch.pth")
state = {"epoch": epoch, "model": model}
torch.save(state, model_out_path)
print("Save best model at {} epoch!!!".format(epoch))
if __name__ == "__main__":
main()