未验证 提交 d77db841 编写于 作者: littletomatodonkey's avatar littletomatodonkey 提交者: GitHub

fix whl exit bug (#3806)

* fix whl exit bug

* fix recognition doc
上级 972bfe07
...@@ -185,11 +185,11 @@ python3 -m paddle.distributed.launch --gpus '0,1,2,3' tools/train.py -c configs ...@@ -185,11 +185,11 @@ python3 -m paddle.distributed.launch --gpus '0,1,2,3' tools/train.py -c configs
<a name="数据增强"></a> <a name="数据增强"></a>
#### 2.1 数据增强 #### 2.1 数据增强
PaddleOCR提供了多种数据增强方式,如果您希望在训练时加入扰动,请在配置文件中设置 `distort: true` PaddleOCR提供了多种数据增强方式,默认配置文件中已经添加了数据增广
默认的扰动方式有:颜色空间转换(cvtColor)、模糊(blur)、抖动(jitter)、噪声(Gasuss noise)、随机切割(random crop)、透视(perspective)、颜色反转(reverse)。 默认的扰动方式有:颜色空间转换(cvtColor)、模糊(blur)、抖动(jitter)、噪声(Gasuss noise)、随机切割(random crop)、透视(perspective)、颜色反转(reverse)、TIA数据增广
训练过程中每种扰动方式以50%的概率被选择,具体代码实现请参考:[img_tools.py](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/ppocr/data/rec/img_tools.py) 训练过程中每种扰动方式以40%的概率被选择,具体代码实现请参考:[rec_img_aug.py](../../ppocr/data/imaug/rec_img_aug.py)
*由于OpenCV的兼容性问题,扰动操作暂时只支持Linux* *由于OpenCV的兼容性问题,扰动操作暂时只支持Linux*
......
...@@ -177,11 +177,11 @@ python3 -m paddle.distributed.launch --gpus '0,1,2,3' tools/train.py -c configs ...@@ -177,11 +177,11 @@ python3 -m paddle.distributed.launch --gpus '0,1,2,3' tools/train.py -c configs
<a name="Data_Augmentation"></a> <a name="Data_Augmentation"></a>
#### 2.1 Data Augmentation #### 2.1 Data Augmentation
PaddleOCR provides a variety of data augmentation methods. If you want to add disturbance during training, please set `distort: true` in the configuration file. PaddleOCR provides a variety of data augmentation methods. All the augmentation methods are enabled by default.
The default perturbation methods are: cvtColor, blur, jitter, Gasuss noise, random crop, perspective, color reverse. The default perturbation methods are: cvtColor, blur, jitter, Gasuss noise, random crop, perspective, color reverse, TIA augmentation.
Each disturbance method is selected with a 50% probability during the training process. For specific code implementation, please refer to: [img_tools.py](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/ppocr/data/rec/img_tools.py) Each disturbance method is selected with a 40% probability during the training process. For specific code implementation, please refer to: [rec_img_aug.py](../../ppocr/data/imaug/rec_img_aug.py)
<a name="Training"></a> <a name="Training"></a>
#### 2.2 Training #### 2.2 Training
......
...@@ -49,14 +49,12 @@ def term_mp(sig_num, frame): ...@@ -49,14 +49,12 @@ def term_mp(sig_num, frame):
os.killpg(pgid, signal.SIGKILL) os.killpg(pgid, signal.SIGKILL)
signal.signal(signal.SIGINT, term_mp)
signal.signal(signal.SIGTERM, term_mp)
def build_dataloader(config, mode, device, logger, seed=None): def build_dataloader(config, mode, device, logger, seed=None):
config = copy.deepcopy(config) config = copy.deepcopy(config)
support_dict = ['SimpleDataSet', 'LMDBDataSet', 'PGDataSet', 'PubTabDataSet'] support_dict = [
'SimpleDataSet', 'LMDBDataSet', 'PGDataSet', 'PubTabDataSet'
]
module_name = config[mode]['dataset']['name'] module_name = config[mode]['dataset']['name']
assert module_name in support_dict, Exception( assert module_name in support_dict, Exception(
'DataSet only support {}'.format(support_dict)) 'DataSet only support {}'.format(support_dict))
...@@ -96,4 +94,8 @@ def build_dataloader(config, mode, device, logger, seed=None): ...@@ -96,4 +94,8 @@ def build_dataloader(config, mode, device, logger, seed=None):
return_list=True, return_list=True,
use_shared_memory=use_shared_memory) use_shared_memory=use_shared_memory)
# support exit using ctrl+c
signal.signal(signal.SIGINT, term_mp)
signal.signal(signal.SIGTERM, term_mp)
return data_loader return data_loader
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册