Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Menu
Open sidebar
wangsen
paddle_dbnet
Commits
24f78b5b
Commit
24f78b5b
authored
Apr 04, 2022
by
littletomatodonkey
Browse files
fix title
parent
2094a40c
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
2 additions
and
2 deletions
+2
-2
doc/doc_ch/finetune.md
doc/doc_ch/finetune.md
+2
-2
No files found.
doc/doc_ch/finetune.md
View file @
24f78b5b
...
...
@@ -94,7 +94,7 @@ PaddleOCR提供的配置文件事在8卡训练(相当于总的batch size是`8*
## 3. 文本识别模型微调
###
2
.1 数据选择
###
3
.1 数据选择
*
数据量:不更换字典的情况下,建议至少准备5000张的文本识别数据集用于模型微调;如果更换了字典(不建议),需要的数量更多。
...
...
@@ -103,7 +103,7 @@ PaddleOCR提供的配置文件事在8卡训练(相当于总的batch size是`8*
*
通用中英文数据:在训练的时候,可以在训练集中添加通用真实数据(如在不更换字典的微调场景中,建议添加LSVT、RCTW、MTWI等真实数据),进一步提升模型的泛化性能。
###
2
.2 模型选择
###
3
.2 模型选择
建议选择PP-OCRv2模型(配置文件:
[
ch_PP-OCRv2_rec_distillation.yml
](
../../configs/rec/ch_PP-OCRv2/ch_PP-OCRv2_rec_distillation.yml
)
,预训练模型:
[
ch_PP-OCRv2_rec_train.tar
](
https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_train.tar
)
)进行微调,其精度与泛化性能是目前提供的最优预训练模型。
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment