Skip to content
体验新版
项目
组织
正在加载...
登录
切换导航
打开侧边栏
秋饼
PaddleOCR
提交
0cc083f6
P
PaddleOCR
项目概览
秋饼
/
PaddleOCR
与 Fork 源项目一致
Fork自
PaddlePaddle / PaddleOCR
通知
1
Star
0
Fork
0
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
0
列表
看板
标记
里程碑
合并请求
0
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
P
PaddleOCR
项目概览
项目概览
详情
发布
仓库
仓库
文件
提交
分支
标签
贡献者
分支图
比较
Issue
0
Issue
0
列表
看板
标记
里程碑
合并请求
0
合并请求
0
Pages
分析
分析
仓库分析
DevOps
Wiki
0
Wiki
成员
成员
收起侧边栏
关闭侧边栏
动态
分支图
创建新Issue
提交
Issue看板
体验新版 GitCode,发现更多精彩内容 >>
提交
0cc083f6
编写于
6月 04, 2020
作者:
T
tink2123
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
mv drop last in det model
上级
b55b8eda
变更
9
隐藏空白更改
内联
并排
Showing
9 changed file
with
10 addition
and
18 deletion
+10
-18
configs/det/det_mv3_db.yml
configs/det/det_mv3_db.yml
+0
-1
configs/det/det_mv3_east.yml
configs/det/det_mv3_east.yml
+0
-1
configs/det/det_r50_vd_db.yml
configs/det/det_r50_vd_db.yml
+0
-1
configs/det/det_r50_vd_east.yml
configs/det/det_r50_vd_east.yml
+0
-1
configs/rec/rec_benchmark_reader.yml
configs/rec/rec_benchmark_reader.yml
+1
-1
configs/rec/rec_icdar15_train.yml
configs/rec/rec_icdar15_train.yml
+4
-4
configs/rec/rec_mv3_tps_bilstm_attn.yml
configs/rec/rec_mv3_tps_bilstm_attn.yml
+3
-3
ppocr/data/det/dataset_traversal.py
ppocr/data/det/dataset_traversal.py
+0
-4
ppocr/modeling/architectures/rec_model.py
ppocr/modeling/architectures/rec_model.py
+2
-2
未找到文件。
configs/det/det_mv3_db.yml
浏览文件 @
0cc083f6
...
...
@@ -9,7 +9,6 @@ Global:
eval_batch_step
:
5000
train_batch_size_per_card
:
16
test_batch_size_per_card
:
16
drop_last
:
false
image_shape
:
[
3
,
640
,
640
]
reader_yml
:
./configs/det/det_db_icdar15_reader.yml
pretrain_weights
:
./pretrain_models/MobileNetV3_large_x0_5_pretrained/
...
...
configs/det/det_mv3_east.yml
浏览文件 @
0cc083f6
...
...
@@ -9,7 +9,6 @@ Global:
eval_batch_step
:
5000
train_batch_size_per_card
:
16
test_batch_size_per_card
:
16
drop_last
:
false
image_shape
:
[
3
,
512
,
512
]
reader_yml
:
./configs/det/det_east_icdar15_reader.yml
pretrain_weights
:
./pretrain_models/MobileNetV3_large_x0_5_pretrained/
...
...
configs/det/det_r50_vd_db.yml
浏览文件 @
0cc083f6
...
...
@@ -10,7 +10,6 @@ Global:
train_batch_size_per_card
:
8
test_batch_size_per_card
:
16
image_shape
:
[
3
,
640
,
640
]
drop_last
:
false
reader_yml
:
./configs/det/det_db_icdar15_reader.yml
pretrain_weights
:
./pretrain_models/ResNet50_vd_ssld_pretrained/
save_res_path
:
./output/det_db/predicts_db.txt
...
...
configs/det/det_r50_vd_east.yml
浏览文件 @
0cc083f6
...
...
@@ -10,7 +10,6 @@ Global:
train_batch_size_per_card
:
8
test_batch_size_per_card
:
16
image_shape
:
[
3
,
512
,
512
]
drop_last
:
false
reader_yml
:
./configs/det/det_east_icdar15_reader.yml
pretrain_weights
:
./pretrain_models/ResNet50_vd_ssld_pretrained/
save_res_path
:
./output/det_east/predicts_east.txt
...
...
configs/rec/rec_benchmark_reader.yml
浏览文件 @
0cc083f6
TrainReader
:
reader_function
:
ppocr.data.rec.dataset_traversal,LMDBReader
num_workers
:
8
lmdb_sets_dir
:
./train_data/data_lmdb_release/
training
/
lmdb_sets_dir
:
./train_data/data_lmdb_release/
evaluation
/
EvalReader
:
reader_function
:
ppocr.data.rec.dataset_traversal,LMDBReader
...
...
configs/rec/rec_icdar15_train.yml
浏览文件 @
0cc083f6
Global
:
algorithm
:
CRNN
use_gpu
:
tru
e
use_gpu
:
fals
e
epoch_num
:
1000
log_smooth_window
:
20
print_batch_step
:
10
save_model_dir
:
./output/rec_CRNN
save_epoch_step
:
300
eval_batch_step
:
500
train_batch_size_per_card
:
2
56
test_batch_size_per_card
:
2
56
train_batch_size_per_card
:
2
test_batch_size_per_card
:
2
image_shape
:
[
3
,
32
,
100
]
max_text_length
:
25
character_type
:
en
loss_type
:
ctc
reader_yml
:
./configs/rec/rec_icdar15_reader.yml
pretrain_weights
:
./pretrain_models/rec_mv3_none_bilstm_ctc/best_accuracy
pretrain_weights
:
checkpoints
:
save_inference_dir
:
infer_img
:
...
...
configs/rec/rec_mv3_tps_bilstm_attn.yml
浏览文件 @
0cc083f6
Global
:
algorithm
:
RARE
use_gpu
:
tru
e
use_gpu
:
fals
e
epoch_num
:
72
log_smooth_window
:
20
print_batch_step
:
10
save_model_dir
:
output/rec_RARE
save_epoch_step
:
3
eval_batch_step
:
2000
train_batch_size_per_card
:
2
56
test_batch_size_per_card
:
2
56
train_batch_size_per_card
:
2
test_batch_size_per_card
:
2
image_shape
:
[
3
,
32
,
100
]
max_text_length
:
25
character_type
:
en
...
...
ppocr/data/det/dataset_traversal.py
浏览文件 @
0cc083f6
...
...
@@ -32,7 +32,6 @@ class TrainReader(object):
self
.
num_workers
=
params
[
'num_workers'
]
self
.
label_file_path
=
params
[
'label_file_path'
]
self
.
batch_size
=
params
[
'train_batch_size_per_card'
]
self
.
drop_last
=
params
[
'drop_last'
]
assert
'process_function'
in
params
,
\
"absence process_function in Reader"
self
.
process
=
create_module
(
params
[
'process_function'
])(
params
)
...
...
@@ -62,9 +61,6 @@ class TrainReader(object):
if
len
(
batch_outs
)
==
self
.
batch_size
:
yield
batch_outs
batch_outs
=
[]
if
not
self
.
drop_last
:
if
len
(
batch_outs
)
!=
0
:
yield
batch_outs
return
batch_iter_reader
...
...
ppocr/modeling/architectures/rec_model.py
浏览文件 @
0cc083f6
...
...
@@ -94,9 +94,9 @@ class RecModel(object):
logger
.
info
(
"WARNRNG!!!
\n
"
"TPS does not support variable shape in chinese!"
"We set
default shape=[3,32,320]
, it may affect the inference effect"
"We set
img_shape to be the same
, it may affect the inference effect"
)
image_shape
[
-
1
]
=
320
image_shape
=
deepcopy
(
self
.
image_shape
)
image
=
fluid
.
data
(
name
=
'image'
,
shape
=
image_shape
,
dtype
=
'float32'
)
labels
=
None
loader
=
None
...
...
编辑
预览
Markdown
is supported
0%
请重试
或
添加新附件
.
添加附件
取消
You are about to add
0
people
to the discussion. Proceed with caution.
先完成此消息的编辑!
取消
想要评论请
注册
或
登录