/
Train.py
158 lines (137 loc) · 5.47 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
from sklearn.metrics import roc_auc_score
import torch
import torch.nn as nn
from torchvision import transforms
from torch.utils.data import DataLoader
from src.model import accident
from src.dataset import DADA
from tqdm import tqdm
import os
from tensorboardX import SummaryWriter
import numpy as np
from src.bert import opt
os.environ['CUDA_VISIBLE_DEVICES']= '0'
transform = transforms.Compose(
[
transforms.ToTensor(),
transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),
]
)
# device = ("cuda" if torch.cuda.is_available() else "cpu")
device = torch.device('cuda:0')
num_epochs =20
batch_size =1
shuffle = True
pin_memory = True
num_workers = 1
rootpath=r''
frame_interval=1
input_shape=[224,224]
seed = 123
np.random.seed(seed)
torch.manual_seed(seed)
device = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu')
train_data=DADA(rootpath , 'training', interval=1,transform=transform)
val_data=DADA(rootpath , 'testing', interval=1,transform=transform)
traindata_loader=DataLoader(dataset=train_data, batch_size=batch_size, shuffle=True,
num_workers=num_workers, pin_memory=True,drop_last=True)
valdata_loader=DataLoader(dataset=val_data, batch_size=batch_size , shuffle=False,
num_workers=num_workers, pin_memory=True,drop_last=True)
def write_scalars(logger, epoch, loss):
logger.add_scalars('train/loss',{'loss':loss}, epoch)
def write_test_scalars(logger, epoch, losses, metrics):
# logger.add_scalars('test/loss',{'loss':loss}, epoch)
logger.add_scalars('test/losses/total_loss',{'Loss': losses}, epoch)
logger.add_scalars('test/accuracy/AP',{'AP':metrics['AP'], 'PR80':metrics['PR80']}, epoch)
logger.add_scalars('test/accuracy/time-to-accident',{'mTTA':metrics['mTTA'], 'TTA_R80':metrics['TTA_R80']}, epoch)
def train():
# the path to save model
model_dir =' '
if not os.path.exists(model_dir):
os.makedirs(model_dir)
logs_dir = '../logs'
if not os.path.exists(logs_dir):
os.makedirs(logs_dir)
logger = SummaryWriter(logs_dir)
h_dim = 256
n_layers = 1
depth=4
adim=opt.adim
heads=opt.heads
num_tokens=opt.num_tokens
c_dim=opt.c_dim
s_dim1=opt.s_dim1
s_dim2=opt.s_dim2
keral=opt.keral
num_class=opt.num_class
# model = AccidentXai(num_classes, x_dim, h_dim, z_dim,n_layers).to(device)
model=accident(h_dim,n_layers,depth,adim,heads,num_tokens,c_dim,s_dim1,s_dim2,keral,num_class).to(device)
# opt1=torch.optim.Adam([
# {'params': model.fusion.parameters(),'lr':1e-3},
# {'params': model.features.parameters(), 'lr': 1e-3},
# {'params': model.deconv.parameters(), 'lr': 1e-3},
# ])
# opt2=torch.optim.Adam([
# {'params': model.fusion.parameters(),'lr':1e-2},
# {'params': model.features.parameters(), 'lr': 1e-3},
# {'params': model.gru_net.parameters(), 'lr': 1e-2}
# ])
# opt1 = torch.optim.Adam([
# {'params': model.fusion.parameters(), 'lr': 1e-6},
# {'params': model.features.parameters(), 'lr': 1e-5},
# {'params': model.deconv.parameters(), 'lr':1e-5}
# ])
#
# opt2 = torch.optim.Adam([
# {'params': model.fusion.parameters(), 'lr':1e-5},
# {'params': model.features.parameters(), 'lr': 1e-5},
# {'params': model.gru_net.parameters(), 'lr': 1e-5},
# ])
opt1 = torch.optim.Adam([
{'params': model.fusion.parameters(), 'lr': 1e-6},
{'params': model.features.parameters(), 'lr': 1e-6},
{'params': model.deconv.parameters(), 'lr': 1e-4},
{'params': model.gru_net.parameters(), 'lr': 1e-5},
])
# for name, param in model.gru_net.named_parameters():
# if 'gru.weight' in name or 'gru.bias' in name:
# param.requires_grad = True
# elif 'dense1' in name or 'dense2' in name:
# param.requires_grad = True
# else:
# param.requires_grad = False
scheduler1 = torch.optim.lr_scheduler.StepLR(opt1, step_size=1, gamma=0.1)
model.train()
for epoch in range(num_epochs):
loop = tqdm(traindata_loader ,total = len(traindata_loader), leave = True)
for imgs, focus, info, label, texts in loop:
labels=label
toa = info[0:, 4].to(device)
loop.set_description(f"Epoch [{epoch+1}/{num_epochs}]")
# print(imgs.shape)
imgs = imgs.to(device)
focus = focus.to(device)
labels= np.array(labels).astype(int)
labels = torch.from_numpy(labels)
labels =labels.to(device)
model.to(device)
loss, outputs = model(imgs ,focus,labels.long(),toa,texts)
opt1.zero_grad()
loss['total_loss'].mean().backward()
torch.nn.utils.clip_grad_norm_(model.parameters(),10)
opt1.step()
loop.set_description(f"Epoch [{epoch+1}/{num_epochs}]")
loop.set_postfix(loss = loss['total_loss'].item())
write_scalars(logger,epoch,loss['total_loss'])
if (epoch+1) % 5 == 0:
scheduler1.step()
#test and evaluate the model
if (epoch+1) % 1==0:
model.train()
# save model
# best_model_file = os.path.join(model_dir, 'best_model.pth')
model_file = os.path.join(model_dir, 'saved_model_%02d.pth'%(epoch))
torch.save(model.state_dict(),model_file)
logger.close()
if __name__ == "__main__":
train()