-
Notifications
You must be signed in to change notification settings - Fork 25
/
config.yaml
83 lines (67 loc) · 2.23 KB
/
config.yaml
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
train:
gpu_id: '0'
backbone: 'resnet18'
pretrained: True
HeadName: 'DB'
in_channels: [64, 128, 256, 512] #[256, 512, 1024, 2048]
inner_channels: 256
k: 50
adaptive: True
start_val_epoch: 1000
n_epoch: 1200
batch_size: 16
use_sr: True
sr_lr: 0.00001
base_lr: 0.002
num_workers: 0
show_step: 5
print_format: 'linux' # linux or windows
restore: True
resume: './checkpoints/DB_resnet18_bs_16_ep_1200/DB.pth.tar'
checkpoints: './checkpoints'
is_icdar2015: True
is_transform: True
is_show: False
train_img_format: '.jpg'
val_img_format: '.jpg'
train_img_dir: '/home/aistudio/work/data/icdar/train_img/'
train_gt_dir: '/home/aistudio/work/data/icdar/train_gt/'
val_img_dir: '/home/aistudio/work/data/icdar/test_img/'
val_gt_dir: '/home/aistudio/work/data/icdar/test_gt/'
radom_angle: [-10, 10]
output_path: './outputs_val'
decay_method: 'e_decay' # e_decay: 指数衰减, s_decay: 指定epoch衰减
schedule: [500,800,1000]
gama: 0.1
test:
gpu_id: '0'
pretrained: False
merge_conv_bn: False
adaptive: False
short_side: 736
thresh: 0.5
box_thresh: 0.6
unclip_ratio: 2
min_size: 3
max_candidates: 1000
is_poly: False
is_icdar2015: True
test_img_format: '.jpg'
test_img_dir: '/home/aistudio/work/data/icdar/test_img/'
test_gt_dir: '/home/aistudio/work/data/icdar/test_gt/'
checkpoints: './checkpoints/DB_resnet18_bs_16_ep_1200/DB.pth.tar'
out_dir: './outputs_test'
pruned:
gpu_id: '0'
scale: [73, 77, 81, 85]
base_num: 8
cut_percent: 0.8
pruned_checkpoints: './pruned/checkpoint/pruned_dict.pth.tar'
checkpoints_dict: './pruned/checkpoint/pruned_dict.dict'
save_checkpoints: './pruned/checkpoint'
checkpoints: './checkpoints/DB_resnet18_bs_16_ep_1200/DB.pth.tar'
finetune_lr: 0.0005
resume: './checkpoints/DB_resnet18_bs_16_ep_1200/DB.pth.tar'
restore: True
n_epoch: 100
start_val_epoch: 40