-
Notifications
You must be signed in to change notification settings - Fork 10
/
train_classify.py
163 lines (134 loc) · 6.68 KB
/
train_classify.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
from torch import optim
import torch
import tqdm
from config import get_config
from solver import Solver
from torch.utils.tensorboard import SummaryWriter
import datetime
import os
import codecs, json
import time
from models.model import ClassifyResNet
from utils.loss import ClassifyLoss
from datasets.steel_dataset import classify_provider
from utils.cal_classify_accuracy import Meter
from utils.set_seed import seed_torch
import pickle
import random
class TrainVal():
def __init__(self, config, fold):
# 加载网络模型
self.model_name = config.model_name
self.model = ClassifyResNet(self.model_name, 4, training=True)
if torch.cuda.is_available():
self.model = torch.nn.DataParallel(self.model)
self.model = self.model.cuda()
# 加载超参数
self.lr = config.lr
self.weight_decay = config.weight_decay
self.epoch = config.epoch
self.fold = fold
# 实例化实现各种子函数的 solver 类
self.solver = Solver(self.model)
# 加载损失函数
self.criterion = ClassifyLoss()
# 创建保存权重的路径
self.model_path = os.path.join(config.save_path, config.model_name)
if not os.path.exists(self.model_path):
os.makedirs(self.model_path)
# 保存json文件和初始化tensorboard
TIMESTAMP = "{0:%Y-%m-%dT%H-%M-%S-%d}-classify".format(datetime.datetime.now(), fold)
self.writer = SummaryWriter(log_dir=os.path.join(self.model_path, TIMESTAMP))
with codecs.open(self.model_path + '/'+ TIMESTAMP + '.json', 'w', "utf-8") as json_file:
json.dump({k: v for k, v in config._get_kwargs()}, json_file, ensure_ascii=False)
self.max_accuracy_valid = 0
# 设置随机种子，注意交叉验证部分划分训练集和验证集的时候，要保持种子固定
self.seed = int(time.time())
seed_torch(self.seed)
with open(self.model_path + '/'+ TIMESTAMP + '.pkl','wb') as f:
pickle.dump({'seed': self.seed}, f, -1)
def train(self, train_loader, valid_loader):
''' 完成模型的训练，保存模型与日志
Args:
train_loader: 训练数据的DataLoader
valid_loader: 验证数据的Dataloader
fold: 当前跑的是第几折
'''
optimizer = optim.Adam(self.model.module.parameters(), self.lr, weight_decay=self.weight_decay)
lr_scheduler = optim.lr_scheduler.CosineAnnealingLR(optimizer, self.epoch+10)
global_step = 0
for epoch in range(self.epoch):
epoch += 1
epoch_loss = 0
self.model.train(True)
tbar = tqdm.tqdm(train_loader)
for i, (images, labels) in enumerate(tbar):
# 网络的前向传播与反向传播
labels_predict = self.solver.forward(images)
loss = self.solver.cal_loss(labels, labels_predict, self.criterion)
epoch_loss += loss.item()
self.solver.backword(optimizer, loss)
# 保存到tensorboard，每一步存储一个
self.writer.add_scalar('train_loss', loss.item(), global_step+i)
descript = "Fold: %d, Train Loss: %.7f, lr: %s" % (self.fold, loss.item(), self.lr)
tbar.set_description(desc=descript)
# 每一个epoch完毕之后，执行学习率衰减
lr_scheduler.step()
global_step += len(train_loader)
# Print the log info
print('Finish Epoch [%d/%d], Average Loss: %.7f' % (epoch, self.epoch, epoch_loss/len(tbar)))
# 验证模型
class_neg_accuracy, class_pos_accuracy, class_accuracy, neg_accuracy, pos_accuracy, accuracy, loss_valid = \
self.validation(valid_loader)
if accuracy > self.max_accuracy_valid:
is_best = True
self.max_accuracy_valid = accuracy
else:
is_best = False
state = {
'epoch': epoch,
'state_dict': self.model.module.state_dict(),
'max_accuracy_valid': self.max_accuracy_valid,
}
self.solver.save_checkpoint(os.path.join(self.model_path, '%s_classify_fold%d.pth' % (self.model_name, self.fold)), state, is_best)
self.writer.add_scalar('valid_loss', loss_valid, epoch)
self.writer.add_scalar('valid_accuracy', accuracy, epoch)
self.writer.add_scalar('valid_class_0_accuracy', class_accuracy[0], epoch)
self.writer.add_scalar('valid_class_1_accuracy', class_accuracy[1], epoch)
self.writer.add_scalar('valid_class_2_accuracy', class_accuracy[2], epoch)
self.writer.add_scalar('valid_class_3_accuracy', class_accuracy[3], epoch)
def validation(self, valid_loader):
''' 完成模型的验证过程
Args:
valid_loader: 验证数据的Dataloader
'''
self.model.eval()
meter = Meter()
tbar = tqdm.tqdm(valid_loader)
loss_sum = 0
with torch.no_grad():
for i, (images, labels) in enumerate(tbar):
# 完成网络的前向传播
labels_predict = self.solver.forward(images)
loss = self.solver.cal_loss(labels, labels_predict, self.criterion)
loss_sum += loss.item()
meter.update(labels, labels_predict.cpu())
descript = "Val Loss: {:.7f}".format(loss.item())
tbar.set_description(desc=descript)
loss_mean = loss_sum / len(tbar)
class_neg_accuracy, class_pos_accuracy, class_accuracy, neg_accuracy, pos_accuracy, accuracy = meter.get_metrics()
print("Class_0_accuracy: %0.4f | Class_1_accuracy: %0.4f | Class_2_accuracy: %0.4f | Class_3_accuracy: %0.4f | "
"Negative accuracy: %0.4f | positive accuracy: %0.4f | accuracy: %0.4f" %
(class_accuracy[0], class_accuracy[1], class_accuracy[2], class_accuracy[3],
neg_accuracy, pos_accuracy, accuracy))
return class_neg_accuracy, class_pos_accuracy, class_accuracy, neg_accuracy, pos_accuracy, accuracy, loss_mean
if __name__ == "__main__":
config = get_config()
mean=(0.485, 0.456, 0.406)
std=(0.229, 0.224, 0.225)
dataloaders = classify_provider(config.dataset_root, os.path.join(config.dataset_root, 'train.csv'), mean, std, config.batch_size, config.num_workers, config.n_splits)
for fold_index, [train_loader, valid_loader] in enumerate(dataloaders):
if fold_index == 0:
continue
train_val = TrainVal(config, fold_index)
train_val.train(train_loader, valid_loader)