-
Notifications
You must be signed in to change notification settings - Fork 20
/
svtr_base_nrtr_syn.yml
119 lines (110 loc) · 2.95 KB
/
svtr_base_nrtr_syn.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
113
114
115
116
117
118
119
Global:
device: gpu
epoch_num: 60
log_smooth_window: 20
print_batch_step: 10
output_dir: ./output/rec/syn/svtr_base_nrtr/
save_epoch_step: 1
# evaluation is run every 2000 iterations
eval_batch_step: [0, 500]
eval_epoch_step: [0, 1]
cal_metric_during_train: True
pretrained_model:
checkpoints:
use_tensorboard: false
infer_img:
# for data or label process
character_dict_path: &character_dict_path ./tools/utils/EN_symbol_dict.txt # 96en
# ./tools/utils/ppocr_keys_v1.txt # ch
max_text_length: &max_text_length 25
use_space_char: &use_space_char False
save_res_path: ./output/rec/syn/predicts_svtr_base_nrtr.txt
use_amp: True
Optimizer:
name: AdamW
lr: 0.0005 # for 4gpus bs256/gpu
weight_decay: 0.05
filter_bias_and_bn: True
LRScheduler:
name: CosineAnnealingLR
warmup_epoch: 6
Architecture:
model_type: rec
algorithm: NRTR
in_channels: 3
Transform:
Encoder:
name: SVTRNet
img_size: [32, 100]
out_char_num: 25
out_channels: 256
patch_merging: 'Conv'
embed_dim: [128, 256, 384]
depth: [6, 6, 6]
num_heads: [4, 8, 12]
mixer: ['Conv','Conv','Conv','Conv','Conv','Conv', 'Conv','Conv', 'Global','Global','Global','Global','Global','Global','Global','Global','Global','Global']
local_mixer: [[5, 5], [5, 5], [5, 5]]
last_stage: False
prenorm: True
Decoder:
name: NRTRDecoder
num_encoder_layers: -1
beam_size: 0
num_decoder_layers: 6
nhead: 12
max_len: *max_text_length
Loss:
name: ARLoss
PostProcess:
name: ARLabelDecode
character_dict_path: *character_dict_path
use_space_char: *use_space_char
Metric:
name: RecMetric
main_indicator: acc
is_filter: True
Train:
dataset:
name: STRLMDBDataSet
data_dir: ./
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
# - SVTRRecAug:
- ARLabelEncode: # Class handling label
character_dict_path: *character_dict_path
use_space_char: *use_space_char
max_text_length: *max_text_length
- SVTRResize:
image_shape: [3, 32, 100]
padding: False
- KeepKeys:
keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order
loader:
shuffle: True
batch_size_per_card: 256
drop_last: True
num_workers: 4
Eval:
dataset:
name: LMDBDataSet
data_dir: ../evaluation/
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- ARLabelEncode: # Class handling label
character_dict_path: *character_dict_path
use_space_char: *use_space_char
max_text_length: *max_text_length
- SVTRResize:
image_shape: [3, 32, 100]
padding: False
- KeepKeys:
keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order
loader:
shuffle: False
drop_last: False
batch_size_per_card: 256
num_workers: 4