-
Notifications
You must be signed in to change notification settings - Fork 7.8k
/
rec_resnet_rfl_att.yml
112 lines (104 loc) · 2.52 KB
/
rec_resnet_rfl_att.yml
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
Global:
use_gpu: True
epoch_num: 6
log_smooth_window: 20
print_batch_step: 50
save_model_dir: ./output/rec/rec_resnet_rfl_att/
save_epoch_step: 1
# evaluation is run every 5000 iterations after the 4000th iteration
eval_batch_step: [0, 5000]
cal_metric_during_train: True
pretrained_model: ./pretrain_models/rec_resnet_rfl_visual/best_accuracy.pdparams
checkpoints:
save_inference_dir:
use_visualdl: False
infer_img: doc/imgs_words_en/word_10.png
# for data or label process
character_dict_path:
max_text_length: 25
infer_mode: False
use_space_char: False
save_res_path: ./output/rec/rec_resnet_rfl.txt
Optimizer:
name: AdamW
beta1: 0.9
beta2: 0.999
weight_decay: 0.0
clip_norm_global: 5.0
lr:
name: Piecewise
decay_epochs : [3, 4, 5]
values : [0.001, 0.0003, 0.00009, 0.000027]
Architecture:
model_type: rec
algorithm: RFL
in_channels: 1
Transform:
name: TPS
num_fiducial: 20
loc_lr: 1.0
model_name: large
Backbone:
name: ResNetRFL
use_cnt: True
use_seq: True
Neck:
name: RFAdaptor
use_v2s: True
use_s2v: True
Head:
name: RFLHead
in_channels: 512
hidden_size: 256
batch_max_legnth: 25
out_channels: 38
use_cnt: True
use_seq: True
Loss:
name: RFLLoss
# ignore_index: 0
PostProcess:
name: RFLLabelDecode
Metric:
name: RecMetric
main_indicator: acc
Train:
dataset:
name: LMDBDataSet
data_dir: ./train_data/data_lmdb_release/training
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- RFLLabelEncode: # Class handling label
- RFLRecResizeImg:
image_shape: [1, 32, 100]
padding: false
interpolation: 2
- KeepKeys:
keep_keys: ['image', 'label', 'length', 'cnt_label'] # dataloader will return list in this order
loader:
shuffle: True
batch_size_per_card: 64
drop_last: True
num_workers: 8
Eval:
dataset:
name: LMDBDataSet
data_dir: ./train_data/data_lmdb_release/validation/
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- RFLLabelEncode: # Class handling label
- RFLRecResizeImg:
image_shape: [1, 32, 100]
padding: false
interpolation: 2
- KeepKeys:
keep_keys: ['image', 'label', 'length', 'cnt_label'] # dataloader will return list in this order
loader:
shuffle: False
drop_last: False
batch_size_per_card: 256
num_workers: 8