@@ -100,7 +100,7 @@ Considering that the features of some channels will be suppressed if the convolu
The recognition module of PP-OCRv3 is optimized based on the text recognition algorithm [SVTR](https://arxiv.org/abs/2205.00159). RNN is abandoned in SVTR, and the context information of the text line image is more effectively mined by introducing the Transformers structure, thereby improving the text recognition ability.
-The recognition accuracy of SVTR_inty outperforms PP-OCRv2 recognition model by 5.3%, while the prediction speed nearly 11 times slower. It takes nearly 100ms to predict a text line on CPU. Therefore, as shown in the figure below, PP-OCRv3 adopts the following six optimization strategies to accelerate the recognition model.
+The recognition accuracy of SVTR_tiny outperforms PP-OCRv2 recognition model by 5.3%, while the prediction speed nearly 11 times slower. It takes nearly 100ms to predict a text line on CPU. Therefore, as shown in the figure below, PP-OCRv3 adopts the following six optimization strategies to accelerate the recognition model.
diff --git a/doc/doc_en/algorithm_det_east_en.md b/doc/doc_en/algorithm_det_east_en.md
index 3955809a49..07c434a9b1 100644
--- a/doc/doc_en/algorithm_det_east_en.md
+++ b/doc/doc_en/algorithm_det_east_en.md
@@ -40,7 +40,7 @@ Please prepare your environment referring to [prepare the environment](./environ
The above EAST model is trained using the ICDAR2015 text detection public dataset. For the download of the dataset, please refer to [ocr_datasets](./dataset/ocr_datasets_en.md).
-After the data download is complete, please refer to [Text Detection Training Tutorial](./detection.md) for training. PaddleOCR has modularized the code structure, so that you only need to **replace the configuration file** to train different detection models.
+After the data download is complete, please refer to [Text Detection Training Tutorial](./detection_en.md) for training. PaddleOCR has modularized the code structure, so that you only need to **replace the configuration file** to train different detection models.
diff --git a/doc/doc_en/algorithm_det_fcenet_en.md b/doc/doc_en/algorithm_det_fcenet_en.md
index e15fb9a07e..f3c51a91a4 100644
--- a/doc/doc_en/algorithm_det_fcenet_en.md
+++ b/doc/doc_en/algorithm_det_fcenet_en.md
@@ -37,7 +37,7 @@ Please prepare your environment referring to [prepare the environment](./environ
The above FCE model is trained using the CTW1500 text detection public dataset. For the download of the dataset, please refer to [ocr_datasets](./dataset/ocr_datasets_en.md).
-After the data download is complete, please refer to [Text Detection Training Tutorial](./detection.md) for training. PaddleOCR has modularized the code structure, so that you only need to **replace the configuration file** to train different detection models.
+After the data download is complete, please refer to [Text Detection Training Tutorial](./detection_en.md) for training. PaddleOCR has modularized the code structure, so that you only need to **replace the configuration file** to train different detection models.
## 4. Inference and Deployment
diff --git a/doc/doc_en/algorithm_det_psenet_en.md b/doc/doc_en/algorithm_det_psenet_en.md
index d4cb3ea7d1..3977a156ac 100644
--- a/doc/doc_en/algorithm_det_psenet_en.md
+++ b/doc/doc_en/algorithm_det_psenet_en.md
@@ -39,7 +39,7 @@ Please prepare your environment referring to [prepare the environment](./environ
The above PSE model is trained using the ICDAR2015 text detection public dataset. For the download of the dataset, please refer to [ocr_datasets](./dataset/ocr_datasets_en.md).
-After the data download is complete, please refer to [Text Detection Training Tutorial](./detection.md) for training. PaddleOCR has modularized the code structure, so that you only need to **replace the configuration file** to train different detection models.
+After the data download is complete, please refer to [Text Detection Training Tutorial](./detection_en.md) for training. PaddleOCR has modularized the code structure, so that you only need to **replace the configuration file** to train different detection models.
## 4. Inference and Deployment
diff --git a/doc/doc_en/algorithm_e2e_pgnet_en.md b/doc/doc_en/algorithm_e2e_pgnet_en.md
index c7cb3221cc..ab74c57bc3 100644
--- a/doc/doc_en/algorithm_e2e_pgnet_en.md
+++ b/doc/doc_en/algorithm_e2e_pgnet_en.md
@@ -36,7 +36,7 @@ The results of detection and recognition are as follows:
## 2. Environment Configuration
-Please refer to [Operation Environment Preparation](./environment_en.md) to configure PaddleOCR operating environment first, refer to [PaddleOCR Overview and Project Clone](./paddleOCR_overview_en.md) to clone the project
+Please refer to [Operation Environment Preparation](./environment_en.md) to configure PaddleOCR operating environment first, refer to [Project Clone](./clone_en.md) to clone the project
## 3. Quick Use
diff --git a/doc/doc_en/algorithm_overview_en.md b/doc/doc_en/algorithm_overview_en.md
index 18c9cd7d51..28aca7c0d1 100755
--- a/doc/doc_en/algorithm_overview_en.md
+++ b/doc/doc_en/algorithm_overview_en.md
@@ -41,6 +41,12 @@ On Total-Text dataset, the text detection result is as follows:
| --- | --- | --- | --- | --- | --- |
|SAST|ResNet50_vd|89.63%|78.44%|83.66%|[trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_sast_totaltext_v2.0_train.tar)|
+On CTW1500 dataset, the text detection result is as follows:
+
+|Model|Backbone|Precision|Recall|Hmean| Download link|
+| --- | --- | --- | --- | --- |---|
+|FCE|ResNet50_dcn|88.39%|82.18%|85.27%| [trained model](https://paddleocr.bj.bcebos.com/contribution/det_r50_dcn_fce_ctw_v2.0_train.tar) |
+
**Note：** Additional data, like icdar2013, icdar2017, COCO-Text, ArT, was added to the model training of SAST. Download English public dataset in organized format used by PaddleOCR from:
* [Baidu Drive](https://pan.baidu.com/s/12cPnZcVuV1zn5DOd4mqjVw) (download code: 2bpi).
* [Google Drive](https://drive.google.com/drive/folders/1ll2-XEVyCQLpJjawLDiRlvo_i4BqHCJe?usp=sharing)
@@ -59,6 +65,8 @@ Supported text recognition algorithms (Click the link to get the tutorial):
- [x] [SAR](./algorithm_rec_sar_en.md)
- [x] [SEED](./algorithm_rec_seed_en.md)
- [x] [SVTR](./algorithm_rec_svtr_en.md)
+- [x] [ViTSTR](./algorithm_rec_vitstr_en.md)
+- [x] [ABINet](./algorithm_rec_abinet_en.md)
Refer to [DTRB](https://arxiv.org/abs/1904.01906), the training and evaluation result of these above text recognition (using MJSynth and SynthText for training, evaluate on IIIT, SVT, IC03, IC13, IC15, SVTP, CUTE) is as follow:
@@ -77,7 +85,8 @@ Refer to [DTRB](https://arxiv.org/abs/1904.01906), the training and evaluation r
|SAR|Resnet31| 87.20% | rec_r31_sar | [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.1/rec/rec_r31_sar_train.tar) |
|SEED|Aster_Resnet| 85.35% | rec_resnet_stn_bilstm_att | [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.1/rec/rec_resnet_stn_bilstm_att.tar) |
|SVTR|SVTR-Tiny| 89.25% | rec_svtr_tiny_none_ctc_en | [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/rec_svtr_tiny_none_ctc_en_train.tar) |
-
+|ViTSTR|ViTSTR| 79.82% | rec_vitstr_none_ce_en | [trained model](https://paddleocr.bj.bcebos.com/rec_vitstr_none_none_train.tar) |
+|ABINet|Resnet45| 90.75% | rec_r45_abinet_en | [trained model](https://paddleocr.bj.bcebos.com/rec_r45_abinet_train.tar) |
diff --git a/doc/doc_en/algorithm_rec_abinet_en.md b/doc/doc_en/algorithm_rec_abinet_en.md
new file mode 100644
index 0000000000..767ca65f64
--- /dev/null
+++ b/doc/doc_en/algorithm_rec_abinet_en.md
@@ -0,0 +1,136 @@
+# ABINet
+
+- [1. Introduction](#1)
+- [2. Environment](#2)
+- [3. Model Training / Evaluation / Prediction](#3)
+ - [3.1 Training](#3-1)
+ - [3.2 Evaluation](#3-2)
+ - [3.3 Prediction](#3-3)
+- [4. Inference and Deployment](#4)
+ - [4.1 Python Inference](#4-1)
+ - [4.2 C++ Inference](#4-2)
+ - [4.3 Serving](#4-3)
+ - [4.4 More](#4-4)
+- [5. FAQ](#5)
+
+
+## 1. Introduction
+
+Paper:
+> [ABINet: Read Like Humans: Autonomous, Bidirectional and Iterative Language Modeling for Scene Text Recognition](https://openaccess.thecvf.com/content/CVPR2021/papers/Fang_Read_Like_Humans_Autonomous_Bidirectional_and_Iterative_Language_Modeling_for_CVPR_2021_paper.pdf)
+> Shancheng Fang and Hongtao Xie and Yuxin Wang and Zhendong Mao and Yongdong Zhang
+> CVPR, 2021
+
+Using MJSynth and SynthText two text recognition datasets for training, and evaluating on IIIT, SVT, IC03, IC13, IC15, SVTP, CUTE datasets, the algorithm reproduction effect is as follows:
+
+|Model|Backbone|config|Acc|Download link|
+| --- | --- | --- | --- | --- |
+|ABINet|ResNet45|[rec_r45_abinet.yml](../../configs/rec/rec_r45_abinet.yml)|90.75%|[pretrained & trained model](https://paddleocr.bj.bcebos.com/rec_r45_abinet_train.tar)|
+
+
+## 2. Environment
+Please refer to ["Environment Preparation"](./environment_en.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone_en.md) to clone the project code.
+
+
+
+## 3. Model Training / Evaluation / Prediction
+
+Please refer to [Text Recognition Tutorial](./recognition_en.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
+
+Training:
+
+Specifically, after the data preparation is completed, the training can be started. The training command is as follows:
+
+```
+#Single GPU training (long training period, not recommended)
+python3 tools/train.py -c configs/rec/rec_r45_abinet.yml
+
+#Multi GPU training, specify the gpu number through the --gpus parameter
+python3 -m paddle.distributed.launch --gpus '0,1,2,3' tools/train.py -c configs/rec/rec_r45_abinet.yml
+```
+
+Evaluation:
+
+```
+# GPU evaluation
+python3 -m paddle.distributed.launch --gpus '0' tools/eval.py -c configs/rec/rec_r45_abinet.yml -o Global.pretrained_model={path/to/weights}/best_accuracy
+```
+
+Prediction:
+
+```
+# The configuration file used for prediction must match the training
+python3 tools/infer_rec.py -c configs/rec/rec_r45_abinet.yml -o Global.infer_img='./doc/imgs_words_en/word_10.png' Global.pretrained_model=./rec_r45_abinet_train/best_accuracy
+```
+
+
+## 4. Inference and Deployment
+
+
+### 4.1 Python Inference
+First, the model saved during the ABINet text recognition training process is converted into an inference model. ( [Model download link](https://paddleocr.bj.bcebos.com/rec_r45_abinet_train.tar)) ), you can use the following command to convert:
+
+```
+python3 tools/export_model.py -c configs/rec/rec_r45_abinet.yml -o Global.pretrained_model=./rec_r45_abinet_train/best_accuracy Global.save_inference_dir=./inference/rec_r45_abinet
+```
+
+**Note:**
+- If you are training the model on your own dataset and have modified the dictionary file, please pay attention to modify the `character_dict_path` in the configuration file to the modified dictionary file.
+- If you modified the input size during training, please modify the `infer_shape` corresponding to ABINet in the `tools/export_model.py` file.
+
+After the conversion is successful, there are three files in the directory:
+```
+/inference/rec_r45_abinet/
+ ├── inference.pdiparams
+ ├── inference.pdiparams.info
+ └── inference.pdmodel
+```
+
+
+For ABINet text recognition model inference, the following commands can be executed:
+
+```
+python3 tools/infer/predict_rec.py --image_dir='./doc/imgs_words_en/word_10.png' --rec_model_dir='./inference/rec_r45_abinet/' --rec_algorithm='ABINet' --rec_image_shape='3,32,128' --rec_char_dict_path='./ppocr/utils/ic15_dict.txt'
+```
+
+
+
+After executing the command, the prediction result (recognized text and score) of the image above is printed to the screen, an example is as follows:
+The result is as follows:
+```shell
+Predicts of ./doc/imgs_words_en/word_10.png:('pain', 0.9999995231628418)
+```
+
+
+### 4.2 C++ Inference
+
+Not supported
+
+
+### 4.3 Serving
+
+Not supported
+
+
+### 4.4 More
+
+Not supported
+
+
+## 5. FAQ
+
+1. Note that the MJSynth and SynthText datasets come from [ABINet repo](https://github.com/FangShancheng/ABINet).
+2. We use the pre-trained model provided by the ABINet authors for finetune training.
+
+## Citation
+
+```bibtex
+@article{Fang2021ABINet,
+ title = {ABINet: Read Like Humans: Autonomous, Bidirectional and Iterative Language Modeling for Scene Text Recognition},
+ author = {Shancheng Fang and Hongtao Xie and Yuxin Wang and Zhendong Mao and Yongdong Zhang},
+ booktitle = {CVPR},
+ year = {2021},
+ url = {https://arxiv.org/abs/2103.06495},
+ pages = {7098-7107}
+}
+```
diff --git a/doc/doc_en/algorithm_rec_aster_en.md b/doc/doc_en/algorithm_rec_aster_en.md
index 1540681a19..b949cb5b37 100644
--- a/doc/doc_en/algorithm_rec_aster_en.md
+++ b/doc/doc_en/algorithm_rec_aster_en.md
@@ -33,13 +33,13 @@ Using MJSynth and SynthText two text recognition datasets for training, and eval
## 2. Environment
-Please refer to ["Environment Preparation"](./environment.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone.md) to clone the project code.
+Please refer to ["Environment Preparation"](./environment_en.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone_en.md) to clone the project code.
## 3. Model Training / Evaluation / Prediction
-Please refer to [Text Recognition Tutorial](./recognition.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
+Please refer to [Text Recognition Tutorial](./recognition_en.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
Training:
diff --git a/doc/doc_en/algorithm_rec_crnn_en.md b/doc/doc_en/algorithm_rec_crnn_en.md
index 571569ee44..8548c2fa62 100644
--- a/doc/doc_en/algorithm_rec_crnn_en.md
+++ b/doc/doc_en/algorithm_rec_crnn_en.md
@@ -33,13 +33,13 @@ Using MJSynth and SynthText two text recognition datasets for training, and eval
## 2. Environment
-Please refer to ["Environment Preparation"](./environment.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone.md) to clone the project code.
+Please refer to ["Environment Preparation"](./environment_en.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone_en.md) to clone the project code.
## 3. Model Training / Evaluation / Prediction
-Please refer to [Text Recognition Tutorial](./recognition.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
+Please refer to [Text Recognition Tutorial](./recognition_en.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
Training:
diff --git a/doc/doc_en/algorithm_rec_nrtr_en.md b/doc/doc_en/algorithm_rec_nrtr_en.md
index 3f8fd0adee..309d7ab123 100644
--- a/doc/doc_en/algorithm_rec_nrtr_en.md
+++ b/doc/doc_en/algorithm_rec_nrtr_en.md
@@ -12,6 +12,7 @@
- [4.3 Serving](#4-3)
- [4.4 More](#4-4)
- [5. FAQ](#5)
+- [6. Release Note](#6)
## 1. Introduction
@@ -25,17 +26,17 @@ Using MJSynth and SynthText two text recognition datasets for training, and eval
|Model|Backbone|config|Acc|Download link|
| --- | --- | --- | --- | --- |
-|NRTR|MTB|[rec_mtb_nrtr.yml](../../configs/rec/rec_mtb_nrtr.yml)|84.21%|[train model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/rec_mtb_nrtr_train.tar)|
+|NRTR|MTB|[rec_mtb_nrtr.yml](../../configs/rec/rec_mtb_nrtr.yml)|84.21%|[trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/rec_mtb_nrtr_train.tar)|
## 2. Environment
-Please refer to ["Environment Preparation"](./environment.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone.md) to clone the project code.
+Please refer to ["Environment Preparation"](./environment_en.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone_en.md) to clone the project code.
## 3. Model Training / Evaluation / Prediction
-Please refer to [Text Recognition Tutorial](./recognition.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
+Please refer to [Text Recognition Tutorial](./recognition_en.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
Training:
@@ -98,7 +99,7 @@ python3 tools/infer/predict_rec.py --image_dir='./doc/imgs_words_en/word_10.png'
After executing the command, the prediction result (recognized text and score) of the image above is printed to the screen, an example is as follows:
The result is as follows:
```shell
-Predicts of ./doc/imgs_words_en/word_10.png:('pain', 0.9265879392623901)
+Predicts of ./doc/imgs_words_en/word_10.png:('pain', 0.9465042352676392)
```
@@ -121,12 +122,146 @@ Not supported
1. In the `NRTR` paper, Beam search is used to decode characters, but the speed is slow. Beam search is not used by default here, and greedy search is used to decode characters.
+
+## 6. Release Note
+
+1. The release/2.6 version updates the NRTR code structure. The new version of NRTR can load the model parameters of the old version (release/2.5 and before), and you may use the following code to convert the old version model parameters to the new version model parameters:
+
+```python
+
+ params = paddle.load('path/' + '.pdparams') # the old version parameters
+ state_dict = model.state_dict() # the new version model parameters
+ new_state_dict = {}
+
+ for k1, v1 in state_dict.items():
+
+ k = k1
+ if 'encoder' in k and 'self_attn' in k and 'qkv' in k and 'weight' in k:
+
+ k_para = k[:13] + 'layers.' + k[13:]
+ q = params[k_para.replace('qkv', 'conv1')].transpose((1, 0, 2, 3))
+ k = params[k_para.replace('qkv', 'conv2')].transpose((1, 0, 2, 3))
+ v = params[k_para.replace('qkv', 'conv3')].transpose((1, 0, 2, 3))
+
+ new_state_dict[k1] = np.concatenate([q[:, :, 0, 0], k[:, :, 0, 0], v[:, :, 0, 0]], -1)
+
+ elif 'encoder' in k and 'self_attn' in k and 'qkv' in k and 'bias' in k:
+
+ k_para = k[:13] + 'layers.' + k[13:]
+ q = params[k_para.replace('qkv', 'conv1')]
+ k = params[k_para.replace('qkv', 'conv2')]
+ v = params[k_para.replace('qkv', 'conv3')]
+
+ new_state_dict[k1] = np.concatenate([q, k, v], -1)
+
+ elif 'encoder' in k and 'self_attn' in k and 'out_proj' in k:
+
+ k_para = k[:13] + 'layers.' + k[13:]
+ new_state_dict[k1] = params[k_para]
+
+ elif 'encoder' in k and 'norm3' in k:
+ k_para = k[:13] + 'layers.' + k[13:]
+ new_state_dict[k1] = params[k_para.replace('norm3', 'norm2')]
+
+ elif 'encoder' in k and 'norm1' in k:
+ k_para = k[:13] + 'layers.' + k[13:]
+ new_state_dict[k1] = params[k_para]
+
+
+ elif 'decoder' in k and 'self_attn' in k and 'qkv' in k and 'weight' in k:
+ k_para = k[:13] + 'layers.' + k[13:]
+ q = params[k_para.replace('qkv', 'conv1')].transpose((1, 0, 2, 3))
+ k = params[k_para.replace('qkv', 'conv2')].transpose((1, 0, 2, 3))
+ v = params[k_para.replace('qkv', 'conv3')].transpose((1, 0, 2, 3))
+ new_state_dict[k1] = np.concatenate([q[:, :, 0, 0], k[:, :, 0, 0], v[:, :, 0, 0]], -1)
+
+ elif 'decoder' in k and 'self_attn' in k and 'qkv' in k and 'bias' in k:
+ k_para = k[:13] + 'layers.' + k[13:]
+ q = params[k_para.replace('qkv', 'conv1')]
+ k = params[k_para.replace('qkv', 'conv2')]
+ v = params[k_para.replace('qkv', 'conv3')]
+ new_state_dict[k1] = np.concatenate([q, k, v], -1)
+
+ elif 'decoder' in k and 'self_attn' in k and 'out_proj' in k:
+
+ k_para = k[:13] + 'layers.' + k[13:]
+ new_state_dict[k1] = params[k_para]
+
+ elif 'decoder' in k and 'cross_attn' in k and 'q' in k and 'weight' in k:
+ k_para = k[:13] + 'layers.' + k[13:]
+ k_para = k_para.replace('cross_attn', 'multihead_attn')
+ q = params[k_para.replace('q', 'conv1')].transpose((1, 0, 2, 3))
+ new_state_dict[k1] = q[:, :, 0, 0]
+
+ elif 'decoder' in k and 'cross_attn' in k and 'q' in k and 'bias' in k:
+ k_para = k[:13] + 'layers.' + k[13:]
+ k_para = k_para.replace('cross_attn', 'multihead_attn')
+ q = params[k_para.replace('q', 'conv1')]
+ new_state_dict[k1] = q
+
+ elif 'decoder' in k and 'cross_attn' in k and 'kv' in k and 'weight' in k:
+ k_para = k[:13] + 'layers.' + k[13:]
+ k_para = k_para.replace('cross_attn', 'multihead_attn')
+ k = params[k_para.replace('kv', 'conv2')].transpose((1, 0, 2, 3))
+ v = params[k_para.replace('kv', 'conv3')].transpose((1, 0, 2, 3))
+ new_state_dict[k1] = np.concatenate([k[:, :, 0, 0], v[:, :, 0, 0]], -1)
+
+ elif 'decoder' in k and 'cross_attn' in k and 'kv' in k and 'bias' in k:
+ k_para = k[:13] + 'layers.' + k[13:]
+ k_para = k_para.replace('cross_attn', 'multihead_attn')
+ k = params[k_para.replace('kv', 'conv2')]
+ v = params[k_para.replace('kv', 'conv3')]
+ new_state_dict[k1] = np.concatenate([k, v], -1)
+
+ elif 'decoder' in k and 'cross_attn' in k and 'out_proj' in k:
+
+ k_para = k[:13] + 'layers.' + k[13:]
+ k_para = k_para.replace('cross_attn', 'multihead_attn')
+ new_state_dict[k1] = params[k_para]
+ elif 'decoder' in k and 'norm' in k:
+ k_para = k[:13] + 'layers.' + k[13:]
+ new_state_dict[k1] = params[k_para]
+ elif 'mlp' in k and 'weight' in k:
+ k_para = k[:13] + 'layers.' + k[13:]
+ k_para = k_para.replace('fc', 'conv')
+ k_para = k_para.replace('mlp.', '')
+ w = params[k_para].transpose((1, 0, 2, 3))
+ new_state_dict[k1] = w[:, :, 0, 0]
+ elif 'mlp' in k and 'bias' in k:
+ k_para = k[:13] + 'layers.' + k[13:]
+ k_para = k_para.replace('fc', 'conv')
+ k_para = k_para.replace('mlp.', '')
+ w = params[k_para]
+ new_state_dict[k1] = w
+
+ else:
+ new_state_dict[k1] = params[k1]
+
+ if list(new_state_dict[k1].shape) != list(v1.shape):
+ print(k1)
+
+
+ for k, v1 in state_dict.items():
+ if k not in new_state_dict.keys():
+ print(1, k)
+ elif list(new_state_dict[k].shape) != list(v1.shape):
+ print(2, k)
+
+
+
+ model.set_state_dict(new_state_dict)
+ paddle.save(model.state_dict(), 'nrtrnew_from_old_params.pdparams')
+
+```
+
+2. The new version has a clean code structure and improved inference speed compared with the old version.
+
## Citation
```bibtex
@article{Sheng2019NRTR,
title = {NRTR: A No-Recurrence Sequence-to-Sequence Model For Scene Text Recognition},
- author = {Fenfen Sheng and Zhineng Chen andBo Xu},
+ author = {Fenfen Sheng and Zhineng Chen and Bo Xu},
booktitle = {ICDAR},
year = {2019},
url = {http://arxiv.org/abs/1806.00926},
diff --git a/doc/doc_en/algorithm_rec_sar_en.md b/doc/doc_en/algorithm_rec_sar_en.md
index 8c1e6dbbfa..24b87c10c3 100644
--- a/doc/doc_en/algorithm_rec_sar_en.md
+++ b/doc/doc_en/algorithm_rec_sar_en.md
@@ -31,13 +31,13 @@ Note:In addition to using the two text recognition datasets MJSynth and SynthTex
## 2. Environment
-Please refer to ["Environment Preparation"](./environment.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone.md) to clone the project code.
+Please refer to ["Environment Preparation"](./environment_en.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone_en.md) to clone the project code.
## 3. Model Training / Evaluation / Prediction
-Please refer to [Text Recognition Tutorial](./recognition.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
+Please refer to [Text Recognition Tutorial](./recognition_en.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
Training:
diff --git a/doc/doc_en/algorithm_rec_seed_en.md b/doc/doc_en/algorithm_rec_seed_en.md
index 21679f42fd..f8d7ae6d3f 100644
--- a/doc/doc_en/algorithm_rec_seed_en.md
+++ b/doc/doc_en/algorithm_rec_seed_en.md
@@ -31,13 +31,13 @@ Using MJSynth and SynthText two text recognition datasets for training, and eval
## 2. Environment
-Please refer to ["Environment Preparation"](./environment.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone.md) to clone the project code.
+Please refer to ["Environment Preparation"](./environment_en.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone_en.md) to clone the project code.
## 3. Model Training / Evaluation / Prediction
-Please refer to [Text Recognition Tutorial](./recognition.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
+Please refer to [Text Recognition Tutorial](./recognition_en.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
Training:
diff --git a/doc/doc_en/algorithm_rec_srn_en.md b/doc/doc_en/algorithm_rec_srn_en.md
index c022a81f9e..1d7fc07dc2 100644
--- a/doc/doc_en/algorithm_rec_srn_en.md
+++ b/doc/doc_en/algorithm_rec_srn_en.md
@@ -30,13 +30,13 @@ Using MJSynth and SynthText two text recognition datasets for training, and eval
## 2. Environment
-Please refer to ["Environment Preparation"](./environment.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone.md) to clone the project code.
+Please refer to ["Environment Preparation"](./environment_en.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone_en.md) to clone the project code.
## 3. Model Training / Evaluation / Prediction
-Please refer to [Text Recognition Tutorial](./recognition.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
+Please refer to [Text Recognition Tutorial](./recognition_en.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
Training:
diff --git a/doc/doc_en/algorithm_rec_starnet.md b/doc/doc_en/algorithm_rec_starnet.md
new file mode 100644
index 0000000000..dbb53a9c73
--- /dev/null
+++ b/doc/doc_en/algorithm_rec_starnet.md
@@ -0,0 +1,139 @@
+# STAR-Net
+
+- [1. Introduction](#1)
+- [2. Environment](#2)
+- [3. Model Training / Evaluation / Prediction](#3)
+ - [3.1 Training](#3-1)
+ - [3.2 Evaluation](#3-2)
+ - [3.3 Prediction](#3-3)
+- [4. Inference and Deployment](#4)
+ - [4.1 Python Inference](#4-1)
+ - [4.2 C++ Inference](#4-2)
+ - [4.3 Serving](#4-3)
+ - [4.4 More](#4-4)
+- [5. FAQ](#5)
+
+
+## 1. Introduction
+
+Paper information:
+> [STAR-Net: a spatial attention residue network for scene text recognition.](http://www.bmva.org/bmvc/2016/papers/paper043/paper043.pdf)
+> Wei Liu, Chaofeng Chen, Kwan-Yee K. Wong, Zhizhong Su and Junyu Han.
+> BMVC, pages 43.1-43.13, 2016
+
+Refer to [DTRB](https://arxiv.org/abs/1904.01906) text Recognition Training and Evaluation Process . Using MJSynth and SynthText two text recognition datasets for training, and evaluating on IIIT, SVT, IC03, IC13, IC15, SVTP, CUTE datasets, the algorithm reproduction effect is as follows:
+
+|Models|Backbone Networks|Avg Accuracy|Configuration Files|Download Links|
+| --- | --- | --- | --- | --- |
+|StarNet|Resnet34_vd|84.44%|[configs/rec/rec_r34_vd_tps_bilstm_ctc.yml](../../configs/rec/rec_r34_vd_tps_bilstm_ctc.yml)|[trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/rec_r34_vd_tps_bilstm_ctc_v2.0_train.tar)|
+|StarNet|MobileNetV3|81.42%|[configs/rec/rec_mv3_tps_bilstm_ctc.yml](../../configs/rec/rec_mv3_tps_bilstm_ctc.yml)|[ trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/rec_mv3_tps_bilstm_ctc_v2.0_train.tar)|
+
+
+
+## 2. Environment
+Please refer to [Operating Environment Preparation](./environment_en.md) to configure the PaddleOCR operating environment, and refer to [Project Clone](./clone_en.md) to clone the project code.
+
+
+## 3. Model Training / Evaluation / Prediction
+
+Please refer to [Text Recognition Training Tutorial](./recognition_en.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**. Take the backbone network based on Resnet34_vd as an example:
+
+
+### 3.1 Training
+After the data preparation is complete, the training can be started. The training command is as follows:
+
+````
+#Single card training (long training period, not recommended)
+python3 tools/train.py -c configs/rec/rec_r34_vd_tps_bilstm_ctc.yml #Multi-card training, specify the card number through the --gpus parameter
+python3 -m paddle.distributed.launch --gpus '0,1,2,3' tools/train.py -c rec_r34_vd_tps_bilstm_ctc.yml
+ ````
+
+
+### 3.2 Evaluation
+
+````
+# GPU evaluation, Global.pretrained_model is the model to be evaluated
+python3 -m paddle.distributed.launch --gpus '0' tools/eval.py -c configs/rec/rec_r34_vd_tps_bilstm_ctc.yml -o Global.pretrained_model={path/to/weights}/best_accuracy
+ ````
+
+
+### 3.3 Prediction
+
+````
+# The configuration file used for prediction must match the training
+python3 tools/infer_rec.py -c configs/rec/rec_r34_vd_tps_bilstm_ctc.yml -o Global.pretrained_model={path/to/weights}/best_accuracy Global.infer_img=doc/imgs_words/en/word_1.png
+ ````
+
+
+## 4. Inference
+
+
+### 4.1 Python Inference
+First, convert the model saved during the STAR-Net text recognition training process into an inference model. Take the model trained on the MJSynth and SynthText text recognition datasets based on the Resnet34_vd backbone network as an example [Model download address]( https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/rec_r34_vd_none_bilstm_ctc_v2.0_train.tar) , which can be converted using the following command:
+
+```shell
+python3 tools/export_model.py -c configs/rec/rec_r34_vd_tps_bilstm_ctc.yml -o Global.pretrained_model=./rec_r34_vd_tps_bilstm_ctc_v2.0_train/best_accuracy Global.save_inference_dir=./inference/rec_starnet
+ ````
+
+STAR-Net text recognition model inference, you can execute the following commands:
+
+```shell
+python3 tools/infer/predict_rec.py --image_dir="./doc/imgs_words_en/word_336.png" --rec_model_dir="./inference/rec_starnet/" --rec_image_shape="3, 32, 100" --rec_char_dict_path="./ppocr/utils/ic15_dict.txt"
+ ````
+
+
+
+The inference results are as follows:
+
+
+```bash
+Predicts of ./doc/imgs_words_en/word_336.png:('super', 0.9999073)
+```
+
+**Attention** Since the above model refers to the [DTRB](https://arxiv.org/abs/1904.01906) text recognition training and evaluation process, it is different from the ultra-lightweight Chinese recognition model training in two aspects:
+
+- The image resolutions used during training are different. The image resolutions used for training the above models are [3, 32, 100], while for Chinese model training, in order to ensure the recognition effect of long texts, the image resolutions used during training are [ 3, 32, 320]. The default shape parameter of the predictive inference program is the image resolution used for training Chinese, i.e. [3, 32, 320]. Therefore, when inferring the above English model here, it is necessary to set the shape of the recognized image through the parameter rec_image_shape.
+
+- Character list, the experiment in the DTRB paper is only for 26 lowercase English letters and 10 numbers, a total of 36 characters. All uppercase and lowercase characters are converted to lowercase characters, and characters not listed above are ignored and considered spaces. Therefore, there is no input character dictionary here, but a dictionary is generated by the following command. Therefore, the parameter rec_char_dict_path needs to be set during inference, which is specified as an English dictionary "./ppocr/utils/ic15_dict.txt".
+
+```
+self.character_str = "0123456789abcdefghijklmnopqrstuvwxyz"
+dict_character = list(self.character_str)
+
+
+ ```
+
+
+### 4.2 C++ Inference
+
+After preparing the inference model, refer to the [cpp infer](../../deploy/cpp_infer/) tutorial to operate.
+
+
+### 4.3 Serving
+
+After preparing the inference model, refer to the [pdserving](../../deploy/pdserving/) tutorial for Serving deployment, including two modes: Python Serving and C++ Serving.
+
+
+### 4.4 More
+
+The STAR-Net model also supports the following inference deployment methods:
+
+- Paddle2ONNX Inference: After preparing the inference model, refer to the [paddle2onnx](../../deploy/paddle2onnx/) tutorial.
+
+
+## 5. FAQ
+
+## Quote
+
+```bibtex
+@inproceedings{liu2016star,
+ title={STAR-Net: a spatial attention residue network for scene text recognition.},
+ author={Liu, Wei and Chen, Chaofeng and Wong, Kwan-Yee K and Su, Zhizhong and Han, Junyu},
+ booktitle={BMVC},
+ volume={2},
+ pages={7},
+ year={2016}
+}
+```
+
+
diff --git a/doc/doc_en/algorithm_rec_svtr_en.md b/doc/doc_en/algorithm_rec_svtr_en.md
index 2e7deb4c07..37cd35f35a 100644
--- a/doc/doc_en/algorithm_rec_svtr_en.md
+++ b/doc/doc_en/algorithm_rec_svtr_en.md
@@ -34,7 +34,7 @@ The accuracy (%) and model files of SVTR on the public dataset of scene text rec
## 2. Environment
-Please refer to ["Environment Preparation"](./environment.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone.md) to clone the project code.
+Please refer to ["Environment Preparation"](./environment_en.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone_en.md) to clone the project code.
#### Dataset Preparation
@@ -44,7 +44,7 @@ Please refer to ["Environment Preparation"](./environment.md) to configure the P
## 3. Model Training / Evaluation / Prediction
-Please refer to [Text Recognition Tutorial](./recognition.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
+Please refer to [Text Recognition Tutorial](./recognition_en.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
Training:
@@ -88,7 +88,6 @@ python3 tools/export_model.py -c configs/rec/rec_svtrnet.yml -o Global.pretraine
**Note:**
- If you are training the model on your own dataset and have modified the dictionary file, please pay attention to modify the `character_dict_path` in the configuration file to the modified dictionary file.
-- If you modified the input size during training, please modify the `infer_shape` corresponding to SVTR in the `tools/export_model.py` file.
After the conversion is successful, there are three files in the directory:
```
diff --git a/doc/doc_en/algorithm_rec_vitstr_en.md b/doc/doc_en/algorithm_rec_vitstr_en.md
new file mode 100644
index 0000000000..a6f9e2f15d
--- /dev/null
+++ b/doc/doc_en/algorithm_rec_vitstr_en.md
@@ -0,0 +1,134 @@
+# ViTSTR
+
+- [1. Introduction](#1)
+- [2. Environment](#2)
+- [3. Model Training / Evaluation / Prediction](#3)
+ - [3.1 Training](#3-1)
+ - [3.2 Evaluation](#3-2)
+ - [3.3 Prediction](#3-3)
+- [4. Inference and Deployment](#4)
+ - [4.1 Python Inference](#4-1)
+ - [4.2 C++ Inference](#4-2)
+ - [4.3 Serving](#4-3)
+ - [4.4 More](#4-4)
+- [5. FAQ](#5)
+
+
+## 1. Introduction
+
+Paper:
+> [Vision Transformer for Fast and Efficient Scene Text Recognition](https://arxiv.org/abs/2105.08582)
+> Rowel Atienza
+> ICDAR, 2021
+
+Using MJSynth and SynthText two text recognition datasets for training, and evaluating on IIIT, SVT, IC03, IC13, IC15, SVTP, CUTE datasets, the algorithm reproduction effect is as follows:
+
+|Model|Backbone|config|Acc|Download link|
+| --- | --- | --- | --- | --- |
+|ViTSTR|ViTSTR|[rec_vitstr_none_ce.yml](../../configs/rec/rec_vitstr_none_ce.yml)|79.82%|[trained model](https://paddleocr.bj.bcebos.com/rec_vitstr_none_none_train.tar)|
+
+
+## 2. Environment
+Please refer to ["Environment Preparation"](./environment_en.md) to configure the PaddleOCR environment, and refer to ["Project Clone"](./clone_en.md) to clone the project code.
+
+
+
+## 3. Model Training / Evaluation / Prediction
+
+Please refer to [Text Recognition Tutorial](./recognition_en.md). PaddleOCR modularizes the code, and training different recognition models only requires **changing the configuration file**.
+
+Training:
+
+Specifically, after the data preparation is completed, the training can be started. The training command is as follows:
+
+```
+#Single GPU training (long training period, not recommended)
+python3 tools/train.py -c configs/rec/rec_vitstr_none_ce.yml
+
+#Multi GPU training, specify the gpu number through the --gpus parameter
+python3 -m paddle.distributed.launch --gpus '0,1,2,3' tools/train.py -c configs/rec/rec_vitstr_none_ce.yml
+```
+
+Evaluation:
+
+```
+# GPU evaluation
+python3 -m paddle.distributed.launch --gpus '0' tools/eval.py -c configs/rec/rec_vitstr_none_ce.yml -o Global.pretrained_model={path/to/weights}/best_accuracy
+```
+
+Prediction:
+
+```
+# The configuration file used for prediction must match the training
+python3 tools/infer_rec.py -c configs/rec/rec_vitstr_none_ce.yml -o Global.infer_img='./doc/imgs_words_en/word_10.png' Global.pretrained_model=./rec_vitstr_none_ce_train/best_accuracy
+```
+
+
+## 4. Inference and Deployment
+
+
+### 4.1 Python Inference
+First, the model saved during the ViTSTR text recognition training process is converted into an inference model. ( [Model download link](https://paddleocr.bj.bcebos.com/rec_vitstr_none_none_train.tar)) ), you can use the following command to convert:
+
+```
+python3 tools/export_model.py -c configs/rec/rec_vitstr_none_ce.yml -o Global.pretrained_model=./rec_vitstr_none_ce_train/best_accuracy Global.save_inference_dir=./inference/rec_vitstr
+```
+
+**Note:**
+- If you are training the model on your own dataset and have modified the dictionary file, please pay attention to modify the `character_dict_path` in the configuration file to the modified dictionary file.
+- If you modified the input size during training, please modify the `infer_shape` corresponding to ViTSTR in the `tools/export_model.py` file.
+
+After the conversion is successful, there are three files in the directory:
+```
+/inference/rec_vitstr/
+ ├── inference.pdiparams
+ ├── inference.pdiparams.info
+ └── inference.pdmodel
+```
+
+
+For ViTSTR text recognition model inference, the following commands can be executed:
+
+```
+python3 tools/infer/predict_rec.py --image_dir='./doc/imgs_words_en/word_10.png' --rec_model_dir='./inference/rec_vitstr/' --rec_algorithm='ViTSTR' --rec_image_shape='1,224,224' --rec_char_dict_path='./ppocr/utils/EN_symbol_dict.txt'
+```
+
+
+
+After executing the command, the prediction result (recognized text and score) of the image above is printed to the screen, an example is as follows:
+The result is as follows:
+```shell
+Predicts of ./doc/imgs_words_en/word_10.png:('pain', 0.9998350143432617)
+```
+
+
+### 4.2 C++ Inference
+
+Not supported
+
+
+### 4.3 Serving
+
+Not supported
+
+
+### 4.4 More
+
+Not supported
+
+
+## 5. FAQ
+
+1. In the `ViTSTR` paper, using pre-trained weights on ImageNet1k for initial training, we did not use pre-trained weights in training, and the final accuracy did not change or even improved.
+
+## Citation
+
+```bibtex
+@article{Atienza2021ViTSTR,
+ title = {Vision Transformer for Fast and Efficient Scene Text Recognition},
+ author = {Rowel Atienza},
+ booktitle = {ICDAR},
+ year = {2021},
+ url = {https://arxiv.org/abs/2105.08582}
+}
+```
diff --git a/doc/doc_en/detection_en.md b/doc/doc_en/detection_en.md
index 76e0f8509b..f85bf585cb 100644
--- a/doc/doc_en/detection_en.md
+++ b/doc/doc_en/detection_en.md
@@ -159,7 +159,7 @@ python3 -m paddle.distributed.launch --ips="xx.xx.xx.xx,xx.xx.xx.xx" --gpus '0,1
-o Global.pretrained_model=./pretrain_models/MobileNetV3_large_x0_5_pretrained
```
-**Note:** When using multi-machine and multi-gpu training, you need to replace the ips value in the above command with the address of your machine, and the machines need to be able to ping each other. In addition, training needs to be launched separately on multiple machines. The command to view the ip address of the machine is `ifconfig`.
+**Note:** (1) When using multi-machine and multi-gpu training, you need to replace the ips value in the above command with the address of your machine, and the machines need to be able to ping each other. (2) Training needs to be launched separately on multiple machines. The command to view the ip address of the machine is `ifconfig`. (3) For more details about the distributed training speedup ratio, please refer to [Distributed Training Tutorial](./distributed_training_en.md).
### 2.6 Training with knowledge distillation
diff --git a/doc/doc_en/distributed_training.md b/doc/doc_en/distributed_training_en.md
similarity index 70%
rename from doc/doc_en/distributed_training.md
rename to doc/doc_en/distributed_training_en.md
index 2822ee5e4e..5a219ed2b4 100644
--- a/doc/doc_en/distributed_training.md
+++ b/doc/doc_en/distributed_training_en.md
@@ -40,11 +40,17 @@ python3 -m paddle.distributed.launch \
## Performance comparison
-* Based on 26W public recognition dataset (LSVT, rctw, mtwi), training on single 8-card P40 and dual 8-card P40, the final time consumption is as follows.
+* On two 8-card P40 graphics cards, the final time consumption and speedup ratio for public recognition dataset (LSVT, RCTW, MTWI) containing 260k images are as follows.
-| Model | Config file | Number of machines | Number of GPUs per machine | Training time | Recognition acc | Speedup ratio |
-| :-------: | :------------: | :----------------: | :----------------------------: | :------------------: | :--------------: | :-----------: |
-| CRNN | configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml | 1 | 8 | 60h | 66.7% | - |
-| CRNN | configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml | 2 | 8 | 40h | 67.0% | 150% |
-It can be seen that the training time is shortened from 60h to 40h, the speedup ratio can reach 150% (60h / 40h), and the efficiency is 75% (60h / (40h * 2)).
+| Model | Config file | Recognition acc | single 8-card training time | two 8-card training time | Speedup ratio |
+|------|-----|--------|--------|--------|-----|
+| CRNN | [rec_chinese_lite_train_v2.0.yml](../../configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml) | 67.0% | 2.50d | 1.67d | **1.5** |
+
+
+* On four 8-card V100 graphics cards, the final time consumption and speedup ratio for full data are as follows.
+
+
+| Model | Config file | Recognition acc | single 8-card training time | four 8-card training time | Speedup ratio |
+|------|-----|--------|--------|--------|-----|
+| SVTR | [ch_PP-OCRv3_rec_distillation.yml](../../configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml) | 74.0% | 10d | 2.84d | **3.5** |
diff --git a/doc/doc_en/knowledge_distillation_en.md b/doc/doc_en/knowledge_distillation_en.md
index bd36907c98..52725e5c05 100755
--- a/doc/doc_en/knowledge_distillation_en.md
+++ b/doc/doc_en/knowledge_distillation_en.md
@@ -438,10 +438,10 @@ Architecture:
```
If DML is used, that is, the method of two small models learning from each other, the Teacher network structure in the above configuration file needs to be set to the same configuration as the Student model.
-Refer to the configuration file for details. [ch_PP-OCRv3_det_dml.yml](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.4/configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_dml.yml)
+Refer to the configuration file for details. [ch_PP-OCRv3_det_dml.yml](../../configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_dml.yml)
-The following describes the configuration file parameters [ch_PP-OCRv3_det_cml.yml](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.4/configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml):
+The following describes the configuration file parameters [ch_PP-OCRv3_det_cml.yml](../../configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml):
```
Architecture:
diff --git a/doc/doc_en/models_list_en.md b/doc/doc_en/models_list_en.md
index 8e8c1f2fe1..c52f71dfe4 100644
--- a/doc/doc_en/models_list_en.md
+++ b/doc/doc_en/models_list_en.md
@@ -20,7 +20,7 @@ The downloadable models provided by PaddleOCR include `inference model`, `traine
|model type|model format|description|
|--- | --- | --- |
-|inference model|inference.pdmodel、inference.pdiparams|Used for inference based on Paddle inference engine，[detail](./inference_en.md)|
+|inference model|inference.pdmodel、inference.pdiparams|Used for inference based on Paddle inference engine，[detail](./inference_ppocr_en.md)|
|trained model, pre-trained model|\*.pdparams、\*.pdopt、\*.states |The checkpoints model saved in the training process, which stores the parameters of the model, mostly used for model evaluation and continuous training.|
|nb model|\*.nb| Model optimized by Paddle-Lite, which is suitable for mobile-side deployment scenarios (Paddle-Lite is needed for nb model deployment). |
@@ -37,7 +37,7 @@ Relationship of the above models is as follows.
|model name|description|config|model size|download|
| --- | --- | --- | --- | --- |
-|ch_PP-OCRv3_det_slim| [New] slim quantization with distillation lightweight model, supporting Chinese, English, multilingual text detection |[ch_PP-OCRv3_det_cml.yml](../../configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml)| 1.1M |[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_slim_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/ch/ch_PP-OCRv3_det_slim_distill_train.tar) / [nb model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_slim_infer.nb)|
+|ch_PP-OCRv3_det_slim| [New] slim quantization with distillation lightweight model, supporting Chinese, English, multilingual text detection |[ch_PP-OCRv3_det_cml.yml](../../configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml)| 1.1M |[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_slim_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_slim_distill_train.tar) / [nb model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_slim_infer.nb)|
|ch_PP-OCRv3_det| [New] Original lightweight model, supporting Chinese, English, multilingual text detection |[ch_PP-OCRv3_det_cml.yml](../../configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml)| 3.8M |[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_distill_train.tar)|
|ch_PP-OCRv2_det_slim| [New] slim quantization with distillation lightweight model, supporting Chinese, English, multilingual text detection|[ch_PP-OCRv2_det_cml.yml](../../configs/det/ch_PP-OCRv2/ch_PP-OCRv2_det_cml.yml)| 3M |[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_slim_quant_infer.tar)|
|ch_PP-OCRv2_det| [New] Original lightweight model, supporting Chinese, English, multilingual text detection|[ch_PP-OCRv2_det_cml.yml](../../configs/det/ch_PP-OCRv2/ch_PP-OCRv2_det_cml.yml)|3M|[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_distill_train.tar)|
@@ -75,7 +75,7 @@ Relationship of the above models is as follows.
|model name|description|config|model size|download|
| --- | --- | --- | --- | --- |
-|ch_PP-OCRv3_rec_slim | [New] Slim qunatization with distillation lightweight model, supporting Chinese, English text recognition |[ch_PP-OCRv3_rec_distillation.yml](../../configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml)| 4.9M |[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_slim_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/ch/ch_PP-OCRv3_rec_slim_train.tar) / [nb model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_slim_infer.nb) |
+|ch_PP-OCRv3_rec_slim | [New] Slim qunatization with distillation lightweight model, supporting Chinese, English text recognition |[ch_PP-OCRv3_rec_distillation.yml](../../configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml)| 4.9M |[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_slim_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_slim_train.tar) / [nb model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_slim_infer.nb) |
|ch_PP-OCRv3_rec| [New] Original lightweight model, supporting Chinese, English, multilingual text recognition |[ch_PP-OCRv3_rec_distillation.yml](../../configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml)| 12.4M |[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_train.tar) |
|ch_PP-OCRv2_rec_slim| Slim qunatization with distillation lightweight model, supporting Chinese, English text recognition|[ch_PP-OCRv2_rec.yml](../../configs/rec/ch_PP-OCRv2/ch_PP-OCRv2_rec.yml)| 9M |[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_slim_quant_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_slim_quant_train.tar) |
|ch_PP-OCRv2_rec| Original lightweight model, supporting Chinese, English, multilingual text recognition |[ch_PP-OCRv2_rec_distillation.yml](../../configs/rec/ch_PP-OCRv2/ch_PP-OCRv2_rec_distillation.yml)|8.5M|[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_train.tar) |
@@ -91,7 +91,7 @@ Relationship of the above models is as follows.
|model name|description|config|model size|download|
| --- | --- | --- | --- | --- |
-|en_PP-OCRv3_rec_slim | [New] Slim qunatization with distillation lightweight model, supporting english, English text recognition |[en_PP-OCRv3_rec.yml](../../configs/rec/PP-OCRv3/en_PP-OCRv3_rec.yml)| 3.2M |[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/english/PP-OCRv3_rec_slim_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_rec_slim_train.tar) / [nb model](https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_rec_slim_infer.nb) |
+|en_PP-OCRv3_rec_slim | [New] Slim qunatization with distillation lightweight model, supporting english, English text recognition |[en_PP-OCRv3_rec.yml](../../configs/rec/PP-OCRv3/en_PP-OCRv3_rec.yml)| 3.2M |[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_rec_slim_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_rec_slim_train.tar) / [nb model](https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_rec_slim_infer.nb) |
|en_PP-OCRv3_rec| [New] Original lightweight model, supporting english, English, multilingual text recognition |[en_PP-OCRv3_rec.yml](../../configs/rec/PP-OCRv3/en_PP-OCRv3_rec.yml)| 9.6M |[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_rec_train.tar) |
|en_number_mobile_slim_v2.0_rec|Slim pruned and quantized lightweight model, supporting English and number recognition|[rec_en_number_lite_train.yml](../../configs/rec/multi_language/rec_en_number_lite_train.yml)| 2.7M | [inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/en_number_mobile_v2.0_rec_slim_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/en_number_mobile_v2.0_rec_slim_train.tar) |
|en_number_mobile_v2.0_rec|Original lightweight model, supporting English and number recognition|[rec_en_number_lite_train.yml](../../configs/rec/multi_language/rec_en_number_lite_train.yml)|2.6M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/en_number_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/en_number_mobile_v2.0_rec_train.tar) |
@@ -108,7 +108,7 @@ Relationship of the above models is as follows.
| ka_PP-OCRv3_rec | ppocr/utils/dict/ka_dict.txt | Lightweight model for Kannada recognition |[ka_PP-OCRv3_rec.yml](../../configs/rec/PP-OCRv3/multi_language/ka_PP-OCRv3_rec.yml)|9.9M|[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/ka_PP-OCRv3_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/ka_PP-OCRv3_rec_train.tar) |
| ta_PP-OCRv3_rec | ppocr/utils/dict/ta_dict.txt |Lightweight model for Tamil recognition|[ta_PP-OCRv3_rec.yml](../../configs/rec/PP-OCRv3/multi_language/ta_PP-OCRv3_rec.yml)|9.6M|[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/ta_PP-OCRv3_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/ta_PP-OCRv3_rec_train.tar) |
| latin_PP-OCRv3_rec | ppocr/utils/dict/latin_dict.txt | Lightweight model for latin recognition | [latin_PP-OCRv3_rec.yml](../../configs/rec/PP-OCRv3/multi_language/latin_PP-OCRv3_rec.yml) |9.7M|[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/latin_PP-OCRv3_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/latin_PP-OCRv3_rec_train.tar) |
-| arabic_PP-OCRv3_rec | ppocr/utils/dict/arabic_dict.txt | Lightweight model for arabic recognition | [arabic_PP-OCRv3_rec.yml](../../configs/rec/PP-OCRv3/multi_language/rec_arabic_lite_train.yml) |9.6M|[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/arabic_PP-OCRv3_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/arabic_PP-OCRv3_rec_train.tar) |
+| arabic_PP-OCRv3_rec | ppocr/utils/dict/arabic_dict.txt | Lightweight model for arabic recognition | [arabic_PP-OCRv3_rec.yml](../../configs/rec/PP-OCRv3/multi_language/arabic_PP-OCRv3_rec.yml) |9.6M|[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/arabic_PP-OCRv3_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/arabic_PP-OCRv3_rec_train.tar) |
| cyrillic_PP-OCRv3_rec | ppocr/utils/dict/cyrillic_dict.txt | Lightweight model for cyrillic recognition | [cyrillic_PP-OCRv3_rec.yml](../../configs/rec/PP-OCRv3/multi_language/cyrillic_PP-OCRv3_rec.yml) |9.6M|[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/cyrillic_PP-OCRv3_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/cyrillic_PP-OCRv3_rec_train.tar) |
| devanagari_PP-OCRv3_rec | ppocr/utils/dict/devanagari_dict.txt | Lightweight model for devanagari recognition | [devanagari_PP-OCRv3_rec.yml](../../configs/rec/PP-OCRv3/multi_language/devanagari_PP-OCRv3_rec.yml) |9.9M|[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/devanagari_PP-OCRv3_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/devanagari_PP-OCRv3_rec_train.tar) |
diff --git a/doc/doc_en/multi_languages_en.md b/doc/doc_en/multi_languages_en.md
index 4696a3e842..d9cb180f70 100644
--- a/doc/doc_en/multi_languages_en.md
+++ b/doc/doc_en/multi_languages_en.md
@@ -187,10 +187,10 @@ In addition to installing the whl package for quick forecasting,
PPOCR also provides a variety of forecasting deployment methods.
If necessary, you can read related documents:
-- [Python Inference](./inference_en.md)
-- [C++ Inference](../../deploy/cpp_infer/readme_en.md)
+- [Python Inference](./inference_ppocr_en.md)
+- [C++ Inference](../../deploy/cpp_infer/readme.md)
- [Serving](../../deploy/hubserving/readme_en.md)
-- [Mobile](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/deploy/lite/readme_en.md)
+- [Mobile](../../deploy/lite/readme.md)
- [Benchmark](./benchmark_en.md)
diff --git a/doc/doc_en/ppocr_introduction_en.md b/doc/doc_en/ppocr_introduction_en.md
index 8fe6bc683a..d28ccb3529 100644
--- a/doc/doc_en/ppocr_introduction_en.md
+++ b/doc/doc_en/ppocr_introduction_en.md
@@ -29,16 +29,16 @@ PP-OCR pipeline is as follows:
PP-OCR system is in continuous optimization. At present, PP-OCR and PP-OCRv2 have been released:
-PP-OCR adopts 19 effective strategies from 8 aspects including backbone network selection and adjustment, prediction head design, data augmentation, learning rate transformation strategy, regularization parameter selection, pre-training model use, and automatic model tailoring and quantization to optimize and slim down the models of each module (as shown in the green box above). The final results are an ultra-lightweight Chinese and English OCR model with an overall size of 3.5M and a 2.8M English digital OCR model. For more details, please refer to the PP-OCR technical article (https://arxiv.org/abs/2009.09941).
+PP-OCR adopts 19 effective strategies from 8 aspects including backbone network selection and adjustment, prediction head design, data augmentation, learning rate transformation strategy, regularization parameter selection, pre-training model use, and automatic model tailoring and quantization to optimize and slim down the models of each module (as shown in the green box above). The final results are an ultra-lightweight Chinese and English OCR model with an overall size of 3.5M and a 2.8M English digital OCR model. For more details, please refer to [PP-OCR technical report](https://arxiv.org/abs/2009.09941).
#### PP-OCRv2
-On the basis of PP-OCR, PP-OCRv2 is further optimized in five aspects. The detection model adopts CML(Collaborative Mutual Learning) knowledge distillation strategy and CopyPaste data expansion strategy. The recognition model adopts LCNet lightweight backbone network, U-DML knowledge distillation strategy and enhanced CTC loss function improvement (as shown in the red box above), which further improves the inference speed and prediction effect. For more details, please refer to the technical report of PP-OCRv2 (https://arxiv.org/abs/2109.03144).
+On the basis of PP-OCR, PP-OCRv2 is further optimized in five aspects. The detection model adopts CML(Collaborative Mutual Learning) knowledge distillation strategy and CopyPaste data expansion strategy. The recognition model adopts LCNet lightweight backbone network, U-DML knowledge distillation strategy and enhanced CTC loss function improvement (as shown in the red box above), which further improves the inference speed and prediction effect. For more details, please refer to [PP-OCRv2 technical report](https://arxiv.org/abs/2109.03144).
#### PP-OCRv3
PP-OCRv3 upgraded the detection model and recognition model in 9 aspects based on PP-OCRv2:
- PP-OCRv3 detector upgrades the CML(Collaborative Mutual Learning) text detection strategy proposed in PP-OCRv2, and further optimizes the effect of teacher model and student model respectively. In the optimization of teacher model, a pan module with large receptive field named LK-PAN is proposed and the DML distillation strategy is adopted; In the optimization of student model, a FPN module with residual attention mechanism named RSE-FPN is proposed.
-- PP-OCRv3 recognizer is optimized based on text recognition algorithm [SVTR](https://arxiv.org/abs/2205.00159). SVTR no longer adopts RNN by introducing transformers structure, which can mine the context information of text line image more effectively, so as to improve the ability of text recognition. PP-OCRv3 adopts lightweight text recognition network SVTR_LCNet, guided training of CTC loss by attention loss, data augmentation strategy TextConAug, better pre-trained model by self-supervised TextRotNet, UDML(Unified Deep Mutual Learning), and UIM (Unlabeled Images Mining) to accelerate the model and improve the effect.
+- PP-OCRv3 recognizer is optimized based on text recognition algorithm [SVTR](https://arxiv.org/abs/2205.00159). SVTR no longer adopts RNN by introducing transformers structure, which can mine the context information of text line image more effectively, so as to improve the ability of text recognition. PP-OCRv3 adopts lightweight text recognition network SVTR_LCNet, guided training of CTC by attention, data augmentation strategy TextConAug, better pre-trained model by self-supervised TextRotNet, UDML(Unified Deep Mutual Learning), and UIM (Unlabeled Images Mining) to accelerate the model and improve the effect.
PP-OCRv3 pipeline is as follows:
@@ -46,7 +46,7 @@ PP-OCRv3 pipeline is as follows:
-For more details, please refer to [PP-OCRv3 technical report](./PP-OCRv3_introduction_en.md).
+For more details, please refer to [PP-OCRv3 technical report](https://arxiv.org/abs/2206.03001v2).
## 2. Features
diff --git a/doc/doc_en/quickstart_en.md b/doc/doc_en/quickstart_en.md
index d7aeb77730..c678dc4762 100644
--- a/doc/doc_en/quickstart_en.md
+++ b/doc/doc_en/quickstart_en.md
@@ -119,7 +119,18 @@ If you do not use the provided test image, you can replace the following `--imag
['PAIN', 0.9934559464454651]
```
-If you need to use the 2.0 model, please specify the parameter `--ocr_version PP-OCR`, paddleocr uses the PP-OCRv3 model by default(`--ocr_version PP-OCRv3`). More whl package usage can be found in [whl package](./whl_en.md)
+**Version**
+paddleocr uses the PP-OCRv3 model by default(`--ocr_version PP-OCRv3`). If you want to use other versions, you can set the parameter `--ocr_version`, the specific version description is as follows:
+| version name | description |
+| --- | --- |
+| PP-OCRv3 | support Chinese and English detection and recognition, direction classifier, support multilingual recognition |
+| PP-OCRv2 | only supports Chinese and English detection and recognition, direction classifier, multilingual model is not updated |
+| PP-OCR | support Chinese and English detection and recognition, direction classifier, support multilingual recognition |
+
+If you want to add your own trained model, you can add model links and keys in [paddleocr](../../paddleocr.py) and recompile.
+
+More whl package usage can be found in [whl package](./whl_en.md)
+
#### 2.1.2 Multi-language Model
diff --git a/doc/doc_en/recognition_en.md b/doc/doc_en/recognition_en.md
index 60b4a1b26b..7d31b0ffe2 100644
--- a/doc/doc_en/recognition_en.md
+++ b/doc/doc_en/recognition_en.md
@@ -306,7 +306,7 @@ python3 -m paddle.distributed.launch --ips="xx.xx.xx.xx,xx.xx.xx.xx" --gpus '0,1
-o Global.pretrained_model=./pretrain_models/rec_mv3_none_bilstm_ctc_v2.0_train
```
-**Note:** When using multi-machine and multi-gpu training, you need to replace the ips value in the above command with the address of your machine, and the machines need to be able to ping each other. In addition, training needs to be launched separately on multiple machines. The command to view the ip address of the machine is `ifconfig`.
+**Note:** (1) When using multi-machine and multi-gpu training, you need to replace the ips value in the above command with the address of your machine, and the machines need to be able to ping each other. (2) Training needs to be launched separately on multiple machines. The command to view the ip address of the machine is `ifconfig`. (3) For more details about the distributed training speedup ratio, please refer to [Distributed Training Tutorial](./distributed_training_en.md).
### 2.6 Training with Knowledge Distillation
diff --git a/doc/doc_en/update_en.md b/doc/doc_en/update_en.md
index a900219b24..a44dd0d70c 100644
--- a/doc/doc_en/update_en.md
+++ b/doc/doc_en/update_en.md
@@ -1,8 +1,8 @@
# RECENT UPDATES
- 2022.5.9 release PaddleOCR v2.5, including:
- - [PP-OCRv3](./doc/doc_en/ppocr_introduction_en.md#pp-ocrv3): With comparable speed, the effect of Chinese scene is further improved by 5% compared with PP-OCRv2, the effect of English scene is improved by 11%, and the average recognition accuracy of 80 language multilingual models is improved by more than 5%.
- - [PPOCRLabelv2](./PPOCRLabel): Add the annotation function for table recognition task, key information extraction task and irregular text image.
- - Interactive e-book [*"Dive into OCR"*](./doc/doc_en/ocr_book_en.md), covers the cutting-edge theory and code practice of OCR full stack technology.
+ - [PP-OCRv3](./ppocr_introduction_en.md#pp-ocrv3): With comparable speed, the effect of Chinese scene is further improved by 5% compared with PP-OCRv2, the effect of English scene is improved by 11%, and the average recognition accuracy of 80 language multilingual models is improved by more than 5%.
+ - [PPOCRLabelv2](../../PPOCRLabel): Add the annotation function for table recognition task, key information extraction task and irregular text image.
+ - Interactive e-book [*"Dive into OCR"*](./ocr_book_en.md), covers the cutting-edge theory and code practice of OCR full stack technology.
- 2022.5.7 Add support for metric and model logging during training to [Weights & Biases](https://docs.wandb.ai/).
- 2021.12.21 OCR open source online course starts. The lesson starts at 8:30 every night and lasts for ten days. Free registration: https://aistudio.baidu.com/aistudio/course/introduce/25207
- 2021.12.21 release PaddleOCR v2.4, release 1 text detection algorithm (PSENet), 3 text recognition algorithms (NRTR、SEED、SAR), 1 key information extraction algorithm (SDMGR) and 3 DocVQA algorithms (LayoutLM、LayoutLMv2，LayoutXLM).
diff --git a/paddleocr.py b/paddleocr.py
index a1265f79de..470dc60da3 100644
--- a/paddleocr.py
+++ b/paddleocr.py
@@ -154,7 +154,13 @@ MODEL_URLS = {
'https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar',
'dict_path': './ppocr/utils/ppocr_keys_v1.txt'
}
- }
+ },
+ 'cls': {
+ 'ch': {
+ 'url':
+ 'https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar',
+ }
+ },
},
'PP-OCR': {
'det': {
diff --git a/ppocr/data/imaug/__init__.py b/ppocr/data/imaug/__init__.py
index 548832fb0d..63dfda91f8 100644
--- a/ppocr/data/imaug/__init__.py
+++ b/ppocr/data/imaug/__init__.py
@@ -22,8 +22,10 @@ from .make_shrink_map import MakeShrinkMap
from .random_crop_data import EastRandomCropData, RandomCropImgMask
from .make_pse_gt import MakePseGt
+
from .rec_img_aug import RecAug, RecConAug, RecResizeImg, ClsResizeImg, \
- SRNRecResizeImg, NRTRRecResizeImg, SARRecResizeImg, PRENResizeImg
+ SRNRecResizeImg, GrayRecResizeImg, SARRecResizeImg, PRENResizeImg, \
+ ABINetRecResizeImg, SVTRRecResizeImg, ABINetRecAug
from .ssl_img_aug import SSLRotateResize
from .randaugment import RandAugment
from .copy_paste import CopyPaste
diff --git a/ppocr/data/imaug/abinet_aug.py b/ppocr/data/imaug/abinet_aug.py
new file mode 100644
index 0000000000..eefdc75d5a
--- /dev/null
+++ b/ppocr/data/imaug/abinet_aug.py
@@ -0,0 +1,407 @@
+# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+This code is refer from:
+https://github.com/FangShancheng/ABINet/blob/main/transforms.py
+"""
+import math
+import numbers
+import random
+
+import cv2
+import numpy as np
+from paddle.vision.transforms import Compose, ColorJitter
+
+
+def sample_asym(magnitude, size=None):
+ return np.random.beta(1, 4, size) * magnitude
+
+
+def sample_sym(magnitude, size=None):
+ return (np.random.beta(4, 4, size=size) - 0.5) * 2 * magnitude
+
+
+def sample_uniform(low, high, size=None):
+ return np.random.uniform(low, high, size=size)
+
+
+def get_interpolation(type='random'):
+ if type == 'random':
+ choice = [
+ cv2.INTER_NEAREST, cv2.INTER_LINEAR, cv2.INTER_CUBIC, cv2.INTER_AREA
+ ]
+ interpolation = choice[random.randint(0, len(choice) - 1)]
+ elif type == 'nearest':
+ interpolation = cv2.INTER_NEAREST
+ elif type == 'linear':
+ interpolation = cv2.INTER_LINEAR
+ elif type == 'cubic':
+ interpolation = cv2.INTER_CUBIC
+ elif type == 'area':
+ interpolation = cv2.INTER_AREA
+ else:
+ raise TypeError(
+ 'Interpolation types only nearest, linear, cubic, area are supported!'
+ )
+ return interpolation
+
+
+class CVRandomRotation(object):
+ def __init__(self, degrees=15):
+ assert isinstance(degrees,
+ numbers.Number), "degree should be a single number."
+ assert degrees >= 0, "degree must be positive."
+ self.degrees = degrees
+
+ @staticmethod
+ def get_params(degrees):
+ return sample_sym(degrees)
+
+ def __call__(self, img):
+ angle = self.get_params(self.degrees)
+ src_h, src_w = img.shape[:2]
+ M = cv2.getRotationMatrix2D(
+ center=(src_w / 2, src_h / 2), angle=angle, scale=1.0)
+ abs_cos, abs_sin = abs(M[0, 0]), abs(M[0, 1])
+ dst_w = int(src_h * abs_sin + src_w * abs_cos)
+ dst_h = int(src_h * abs_cos + src_w * abs_sin)
+ M[0, 2] += (dst_w - src_w) / 2
+ M[1, 2] += (dst_h - src_h) / 2
+
+ flags = get_interpolation()
+ return cv2.warpAffine(
+ img,
+ M, (dst_w, dst_h),
+ flags=flags,
+ borderMode=cv2.BORDER_REPLICATE)
+
+
+class CVRandomAffine(object):
+ def __init__(self, degrees, translate=None, scale=None, shear=None):
+ assert isinstance(degrees,
+ numbers.Number), "degree should be a single number."
+ assert degrees >= 0, "degree must be positive."
+ self.degrees = degrees
+
+ if translate is not None:
+ assert isinstance(translate, (tuple, list)) and len(translate) == 2, \
+ "translate should be a list or tuple and it must be of length 2."
+ for t in translate:
+ if not (0.0 <= t <= 1.0):
+ raise ValueError(
+ "translation values should be between 0 and 1")
+ self.translate = translate
+
+ if scale is not None:
+ assert isinstance(scale, (tuple, list)) and len(scale) == 2, \
+ "scale should be a list or tuple and it must be of length 2."
+ for s in scale:
+ if s <= 0:
+ raise ValueError("scale values should be positive")
+ self.scale = scale
+
+ if shear is not None:
+ if isinstance(shear, numbers.Number):
+ if shear < 0:
+ raise ValueError(
+ "If shear is a single number, it must be positive.")
+ self.shear = [shear]
+ else:
+ assert isinstance(shear, (tuple, list)) and (len(shear) == 2), \
+ "shear should be a list or tuple and it must be of length 2."
+ self.shear = shear
+ else:
+ self.shear = shear
+
+ def _get_inverse_affine_matrix(self, center, angle, translate, scale,
+ shear):
+ # https://github.com/pytorch/vision/blob/v0.4.0/torchvision/transforms/functional.py#L717
+ from numpy import sin, cos, tan
+
+ if isinstance(shear, numbers.Number):
+ shear = [shear, 0]
+
+ if not isinstance(shear, (tuple, list)) and len(shear) == 2:
+ raise ValueError(
+ "Shear should be a single value or a tuple/list containing " +
+ "two values. Got {}".format(shear))
+
+ rot = math.radians(angle)
+ sx, sy = [math.radians(s) for s in shear]
+
+ cx, cy = center
+ tx, ty = translate
+
+ # RSS without scaling
+ a = cos(rot - sy) / cos(sy)
+ b = -cos(rot - sy) * tan(sx) / cos(sy) - sin(rot)
+ c = sin(rot - sy) / cos(sy)
+ d = -sin(rot - sy) * tan(sx) / cos(sy) + cos(rot)
+
+ # Inverted rotation matrix with scale and shear
+ # det([[a, b], [c, d]]) == 1, since det(rotation) = 1 and det(shear) = 1
+ M = [d, -b, 0, -c, a, 0]
+ M = [x / scale for x in M]
+
+ # Apply inverse of translation and of center translation: RSS^-1 * C^-1 * T^-1
+ M[2] += M[0] * (-cx - tx) + M[1] * (-cy - ty)
+ M[5] += M[3] * (-cx - tx) + M[4] * (-cy - ty)
+
+ # Apply center translation: C * RSS^-1 * C^-1 * T^-1
+ M[2] += cx
+ M[5] += cy
+ return M
+
+ @staticmethod
+ def get_params(degrees, translate, scale_ranges, shears, height):
+ angle = sample_sym(degrees)
+ if translate is not None:
+ max_dx = translate[0] * height
+ max_dy = translate[1] * height
+ translations = (np.round(sample_sym(max_dx)),
+ np.round(sample_sym(max_dy)))
+ else:
+ translations = (0, 0)
+
+ if scale_ranges is not None:
+ scale = sample_uniform(scale_ranges[0], scale_ranges[1])
+ else:
+ scale = 1.0
+
+ if shears is not None:
+ if len(shears) == 1:
+ shear = [sample_sym(shears[0]), 0.]
+ elif len(shears) == 2:
+ shear = [sample_sym(shears[0]), sample_sym(shears[1])]
+ else:
+ shear = 0.0
+
+ return angle, translations, scale, shear
+
+ def __call__(self, img):
+ src_h, src_w = img.shape[:2]
+ angle, translate, scale, shear = self.get_params(
+ self.degrees, self.translate, self.scale, self.shear, src_h)
+
+ M = self._get_inverse_affine_matrix((src_w / 2, src_h / 2), angle,
+ (0, 0), scale, shear)
+ M = np.array(M).reshape(2, 3)
+
+ startpoints = [(0, 0), (src_w - 1, 0), (src_w - 1, src_h - 1),
+ (0, src_h - 1)]
+ project = lambda x, y, a, b, c: int(a * x + b * y + c)
+ endpoints = [(project(x, y, *M[0]), project(x, y, *M[1]))
+ for x, y in startpoints]
+
+ rect = cv2.minAreaRect(np.array(endpoints))
+ bbox = cv2.boxPoints(rect).astype(dtype=np.int)
+ max_x, max_y = bbox[:, 0].max(), bbox[:, 1].max()
+ min_x, min_y = bbox[:, 0].min(), bbox[:, 1].min()
+
+ dst_w = int(max_x - min_x)
+ dst_h = int(max_y - min_y)
+ M[0, 2] += (dst_w - src_w) / 2
+ M[1, 2] += (dst_h - src_h) / 2
+
+ # add translate
+ dst_w += int(abs(translate[0]))
+ dst_h += int(abs(translate[1]))
+ if translate[0] < 0: M[0, 2] += abs(translate[0])
+ if translate[1] < 0: M[1, 2] += abs(translate[1])
+
+ flags = get_interpolation()
+ return cv2.warpAffine(
+ img,
+ M, (dst_w, dst_h),
+ flags=flags,
+ borderMode=cv2.BORDER_REPLICATE)
+
+
+class CVRandomPerspective(object):
+ def __init__(self, distortion=0.5):
+ self.distortion = distortion
+
+ def get_params(self, width, height, distortion):
+ offset_h = sample_asym(
+ distortion * height / 2, size=4).astype(dtype=np.int)
+ offset_w = sample_asym(
+ distortion * width / 2, size=4).astype(dtype=np.int)
+ topleft = (offset_w[0], offset_h[0])
+ topright = (width - 1 - offset_w[1], offset_h[1])
+ botright = (width - 1 - offset_w[2], height - 1 - offset_h[2])
+ botleft = (offset_w[3], height - 1 - offset_h[3])
+
+ startpoints = [(0, 0), (width - 1, 0), (width - 1, height - 1),
+ (0, height - 1)]
+ endpoints = [topleft, topright, botright, botleft]
+ return np.array(
+ startpoints, dtype=np.float32), np.array(
+ endpoints, dtype=np.float32)
+
+ def __call__(self, img):
+ height, width = img.shape[:2]
+ startpoints, endpoints = self.get_params(width, height, self.distortion)
+ M = cv2.getPerspectiveTransform(startpoints, endpoints)
+
+ # TODO: more robust way to crop image
+ rect = cv2.minAreaRect(endpoints)
+ bbox = cv2.boxPoints(rect).astype(dtype=np.int)
+ max_x, max_y = bbox[:, 0].max(), bbox[:, 1].max()
+ min_x, min_y = bbox[:, 0].min(), bbox[:, 1].min()
+ min_x, min_y = max(min_x, 0), max(min_y, 0)
+
+ flags = get_interpolation()
+ img = cv2.warpPerspective(
+ img,
+ M, (max_x, max_y),
+ flags=flags,
+ borderMode=cv2.BORDER_REPLICATE)
+ img = img[min_y:, min_x:]
+ return img
+
+
+class CVRescale(object):
+ def __init__(self, factor=4, base_size=(128, 512)):
+ """ Define image scales using gaussian pyramid and rescale image to target scale.
+
+ Args:
+ factor: the decayed factor from base size, factor=4 keeps target scale by default.
+ base_size: base size the build the bottom layer of pyramid
+ """
+ if isinstance(factor, numbers.Number):
+ self.factor = round(sample_uniform(0, factor))
+ elif isinstance(factor, (tuple, list)) and len(factor) == 2:
+ self.factor = round(sample_uniform(factor[0], factor[1]))
+ else:
+ raise Exception('factor must be number or list with length 2')
+ # assert factor is valid
+ self.base_h, self.base_w = base_size[:2]
+
+ def __call__(self, img):
+ if self.factor == 0: return img
+ src_h, src_w = img.shape[:2]
+ cur_w, cur_h = self.base_w, self.base_h
+ scale_img = cv2.resize(
+ img, (cur_w, cur_h), interpolation=get_interpolation())
+ for _ in range(self.factor):
+ scale_img = cv2.pyrDown(scale_img)
+ scale_img = cv2.resize(
+ scale_img, (src_w, src_h), interpolation=get_interpolation())
+ return scale_img
+
+
+class CVGaussianNoise(object):
+ def __init__(self, mean=0, var=20):
+ self.mean = mean
+ if isinstance(var, numbers.Number):
+ self.var = max(int(sample_asym(var)), 1)
+ elif isinstance(var, (tuple, list)) and len(var) == 2:
+ self.var = int(sample_uniform(var[0], var[1]))
+ else:
+ raise Exception('degree must be number or list with length 2')
+
+ def __call__(self, img):
+ noise = np.random.normal(self.mean, self.var**0.5, img.shape)
+ img = np.clip(img + noise, 0, 255).astype(np.uint8)
+ return img
+
+
+class CVMotionBlur(object):
+ def __init__(self, degrees=12, angle=90):
+ if isinstance(degrees, numbers.Number):
+ self.degree = max(int(sample_asym(degrees)), 1)
+ elif isinstance(degrees, (tuple, list)) and len(degrees) == 2:
+ self.degree = int(sample_uniform(degrees[0], degrees[1]))
+ else:
+ raise Exception('degree must be number or list with length 2')
+ self.angle = sample_uniform(-angle, angle)
+
+ def __call__(self, img):
+ M = cv2.getRotationMatrix2D((self.degree // 2, self.degree // 2),
+ self.angle, 1)
+ motion_blur_kernel = np.zeros((self.degree, self.degree))
+ motion_blur_kernel[self.degree // 2, :] = 1
+ motion_blur_kernel = cv2.warpAffine(motion_blur_kernel, M,
+ (self.degree, self.degree))
+ motion_blur_kernel = motion_blur_kernel / self.degree
+ img = cv2.filter2D(img, -1, motion_blur_kernel)
+ img = np.clip(img, 0, 255).astype(np.uint8)
+ return img
+
+
+class CVGeometry(object):
+ def __init__(self,
+ degrees=15,
+ translate=(0.3, 0.3),
+ scale=(0.5, 2.),
+ shear=(45, 15),
+ distortion=0.5,
+ p=0.5):
+ self.p = p
+ type_p = random.random()
+ if type_p < 0.33:
+ self.transforms = CVRandomRotation(degrees=degrees)
+ elif type_p < 0.66:
+ self.transforms = CVRandomAffine(
+ degrees=degrees, translate=translate, scale=scale, shear=shear)
+ else:
+ self.transforms = CVRandomPerspective(distortion=distortion)
+
+ def __call__(self, img):
+ if random.random() < self.p:
+ return self.transforms(img)
+ else:
+ return img
+
+
+class CVDeterioration(object):
+ def __init__(self, var, degrees, factor, p=0.5):
+ self.p = p
+ transforms = []
+ if var is not None:
+ transforms.append(CVGaussianNoise(var=var))
+ if degrees is not None:
+ transforms.append(CVMotionBlur(degrees=degrees))
+ if factor is not None:
+ transforms.append(CVRescale(factor=factor))
+
+ random.shuffle(transforms)
+ transforms = Compose(transforms)
+ self.transforms = transforms
+
+ def __call__(self, img):
+ if random.random() < self.p:
+
+ return self.transforms(img)
+ else:
+ return img
+
+
+class CVColorJitter(object):
+ def __init__(self,
+ brightness=0.5,
+ contrast=0.5,
+ saturation=0.5,
+ hue=0.1,
+ p=0.5):
+ self.p = p
+ self.transforms = ColorJitter(
+ brightness=brightness,
+ contrast=contrast,
+ saturation=saturation,
+ hue=hue)
+
+ def __call__(self, img):
+ if random.random() < self.p: return self.transforms(img)
+ else: return img
diff --git a/ppocr/data/imaug/fce_targets.py b/ppocr/data/imaug/fce_targets.py
index 4d1903c0a7..8c64276e26 100644
--- a/ppocr/data/imaug/fce_targets.py
+++ b/ppocr/data/imaug/fce_targets.py
@@ -107,17 +107,20 @@ class FCENetTargets:
for i in range(1, n):
current_line_len = i * delta_length
- while current_line_len >= length_cumsum[current_edge_ind + 1]:
+ while current_edge_ind + 1 < len(length_cumsum) and current_line_len >= length_cumsum[current_edge_ind + 1]:
current_edge_ind += 1
+
current_edge_end_shift = current_line_len - length_cumsum[
current_edge_ind]
+
+ if current_edge_ind >= len(length_list):
+ break
end_shift_ratio = current_edge_end_shift / length_list[
current_edge_ind]
current_point = line[current_edge_ind] + (line[current_edge_ind + 1]
- line[current_edge_ind]
) * end_shift_ratio
resampled_line.append(current_point)
-
resampled_line.append(line[-1])
resampled_line = np.array(resampled_line)
@@ -328,6 +331,8 @@ class FCENetTargets:
resampled_top_line, resampled_bot_line = self.resample_sidelines(
top_line, bot_line, self.resample_step)
resampled_bot_line = resampled_bot_line[::-1]
+ if len(resampled_top_line) != len(resampled_bot_line):
+ continue
center_line = (resampled_top_line + resampled_bot_line) / 2
line_head_shrink_len = norm(resampled_top_line[0] -
diff --git a/ppocr/data/imaug/label_ops.py b/ppocr/data/imaug/label_ops.py
index 02a5187dad..312d6dc9ad 100644
--- a/ppocr/data/imaug/label_ops.py
+++ b/ppocr/data/imaug/label_ops.py
@@ -23,7 +23,6 @@ import string
from shapely.geometry import LineString, Point, Polygon
import json
import copy
-
from ppocr.utils.logging import get_logger
@@ -74,9 +73,10 @@ class DetLabelEncode(object):
s = pts.sum(axis=1)
rect[0] = pts[np.argmin(s)]
rect[2] = pts[np.argmax(s)]
- diff = np.diff(pts, axis=1)
- rect[1] = pts[np.argmin(diff)]
- rect[3] = pts[np.argmax(diff)]
+ tmp = np.delete(pts, (np.argmin(s), np.argmax(s)), axis=0)
+ diff = np.diff(np.array(tmp), axis=1)
+ rect[1] = tmp[np.argmin(diff)]
+ rect[3] = tmp[np.argmax(diff)]
return rect
def expand_points_num(self, boxes):
@@ -157,37 +157,6 @@ class BaseRecLabelEncode(object):
return text_list
-class NRTRLabelEncode(BaseRecLabelEncode):
- """ Convert between text-label and text-index """
-
- def __init__(self,
- max_text_length,
- character_dict_path=None,
- use_space_char=False,
- **kwargs):
-
- super(NRTRLabelEncode, self).__init__(
- max_text_length, character_dict_path, use_space_char)
-
- def __call__(self, data):
- text = data['label']
- text = self.encode(text)
- if text is None:
- return None
- if len(text) >= self.max_text_len - 1:
- return None
- data['length'] = np.array(len(text))
- text.insert(0, 2)
- text.append(3)
- text = text + [0] * (self.max_text_len - len(text))
- data['label'] = np.array(text)
- return data
-
- def add_special_char(self, dict_character):
- dict_character = ['blank', '
', '', ''] + dict_character
- return dict_character
-
-
class CTCLabelEncode(BaseRecLabelEncode):
""" Convert between text-label and text-index """
@@ -438,12 +407,14 @@ class KieLabelEncode(object):
texts.append(ann['transcription'])
text_ind = [self.dict[c] for c in text if c in self.dict]
text_inds.append(text_ind)
- if 'label' in anno.keys():
+ if 'label' in ann.keys():
labels.append(ann['label'])
- elif 'key_cls' in anno.keys():
- labels.append(anno['key_cls'])
+ elif 'key_cls' in ann.keys():
+ labels.append(ann['key_cls'])
else:
- raise ValueError("Cannot found 'key_cls' in ann.keys(), please check your training annotation.")
+ raise ValueError(
+ "Cannot found 'key_cls' in ann.keys(), please check your training annotation."
+ )
edges.append(ann.get('edge', 0))
ann_infos = dict(
image=data['image'],
@@ -1044,3 +1015,99 @@ class MultiLabelEncode(BaseRecLabelEncode):
data_out['label_sar'] = sar['label']
data_out['length'] = ctc['length']
return data_out
+
+
+class NRTRLabelEncode(BaseRecLabelEncode):
+ """ Convert between text-label and text-index """
+
+ def __init__(self,
+ max_text_length,
+ character_dict_path=None,
+ use_space_char=False,
+ **kwargs):
+
+ super(NRTRLabelEncode, self).__init__(
+ max_text_length, character_dict_path, use_space_char)
+
+ def __call__(self, data):
+ text = data['label']
+ text = self.encode(text)
+ if text is None:
+ return None
+ if len(text) >= self.max_text_len - 1:
+ return None
+ data['length'] = np.array(len(text))
+ text.insert(0, 2)
+ text.append(3)
+ text = text + [0] * (self.max_text_len - len(text))
+ data['label'] = np.array(text)
+ return data
+
+ def add_special_char(self, dict_character):
+ dict_character = ['blank', '', '', ''] + dict_character
+ return dict_character
+
+
+class ViTSTRLabelEncode(BaseRecLabelEncode):
+ """ Convert between text-label and text-index """
+
+ def __init__(self,
+ max_text_length,
+ character_dict_path=None,
+ use_space_char=False,
+ ignore_index=0,
+ **kwargs):
+
+ super(ViTSTRLabelEncode, self).__init__(
+ max_text_length, character_dict_path, use_space_char)
+ self.ignore_index = ignore_index
+
+ def __call__(self, data):
+ text = data['label']
+ text = self.encode(text)
+ if text is None:
+ return None
+ if len(text) >= self.max_text_len:
+ return None
+ data['length'] = np.array(len(text))
+ text.insert(0, self.ignore_index)
+ text.append(1)
+ text = text + [self.ignore_index] * (self.max_text_len + 2 - len(text))
+ data['label'] = np.array(text)
+ return data
+
+ def add_special_char(self, dict_character):
+ dict_character = ['', ''] + dict_character
+ return dict_character
+
+
+class ABINetLabelEncode(BaseRecLabelEncode):
+ """ Convert between text-label and text-index """
+
+ def __init__(self,
+ max_text_length,
+ character_dict_path=None,
+ use_space_char=False,
+ ignore_index=100,
+ **kwargs):
+
+ super(ABINetLabelEncode, self).__init__(
+ max_text_length, character_dict_path, use_space_char)
+ self.ignore_index = ignore_index
+
+ def __call__(self, data):
+ text = data['label']
+ text = self.encode(text)
+ if text is None:
+ return None
+ if len(text) >= self.max_text_len:
+ return None
+ data['length'] = np.array(len(text))
+ text.append(0)
+ text = text + [self.ignore_index] * (self.max_text_len + 1 - len(text))
+ data['label'] = np.array(text)
+ return data
+
+ def add_special_char(self, dict_character):
+ dict_character = [''] + dict_character
+ return dict_character
diff --git a/ppocr/data/imaug/operators.py b/ppocr/data/imaug/operators.py
index 09736515e7..5397d71ccb 100644
--- a/ppocr/data/imaug/operators.py
+++ b/ppocr/data/imaug/operators.py
@@ -67,39 +67,6 @@ class DecodeImage(object):
return data
-class NRTRDecodeImage(object):
- """ decode image """
-
- def __init__(self, img_mode='RGB', channel_first=False, **kwargs):
- self.img_mode = img_mode
- self.channel_first = channel_first
-
- def __call__(self, data):
- img = data['image']
- if six.PY2:
- assert type(img) is str and len(
- img) > 0, "invalid input 'img' in DecodeImage"
- else:
- assert type(img) is bytes and len(
- img) > 0, "invalid input 'img' in DecodeImage"
- img = np.frombuffer(img, dtype='uint8')
-
- img = cv2.imdecode(img, 1)
-
- if img is None:
- return None
- if self.img_mode == 'GRAY':
- img = cv2.cvtColor(img, cv2.COLOR_GRAY2BGR)
- elif self.img_mode == 'RGB':
- assert img.shape[2] == 3, 'invalid shape of image[%s]' % (img.shape)
- img = img[:, :, ::-1]
- img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
- if self.channel_first:
- img = img.transpose((2, 0, 1))
- data['image'] = img
- return data
-
-
class NormalizeImage(object):
""" normalize image such as substract mean, divide std
"""
diff --git a/ppocr/data/imaug/rec_img_aug.py b/ppocr/data/imaug/rec_img_aug.py
index 7483dffe5b..26773d0a51 100644
--- a/ppocr/data/imaug/rec_img_aug.py
+++ b/ppocr/data/imaug/rec_img_aug.py
@@ -19,16 +19,109 @@ import random
import copy
from PIL import Image
from .text_image_aug import tia_perspective, tia_stretch, tia_distort
+from .abinet_aug import CVGeometry, CVDeterioration, CVColorJitter
+from paddle.vision.transforms import Compose
class RecAug(object):
- def __init__(self, use_tia=True, aug_prob=0.4, **kwargs):
- self.use_tia = use_tia
- self.aug_prob = aug_prob
+ def __init__(self,
+ tia_prob=0.4,
+ crop_prob=0.4,
+ reverse_prob=0.4,
+ noise_prob=0.4,
+ jitter_prob=0.4,
+ blur_prob=0.4,
+ hsv_aug_prob=0.4,
+ **kwargs):
+ self.tia_prob = tia_prob
+ self.bda = BaseDataAugmentation(crop_prob, reverse_prob, noise_prob,
+ jitter_prob, blur_prob, hsv_aug_prob)
def __call__(self, data):
img = data['image']
- img = warp(img, 10, self.use_tia, self.aug_prob)
+ h, w, _ = img.shape
+
+ # tia
+ if random.random() <= self.tia_prob:
+ if h >= 20 and w >= 20:
+ img = tia_distort(img, random.randint(3, 6))
+ img = tia_stretch(img, random.randint(3, 6))
+ img = tia_perspective(img)
+
+ # bda
+ data['image'] = img
+ data = self.bda(data)
+ return data
+
+
+class BaseDataAugmentation(object):
+ def __init__(self,
+ crop_prob=0.4,
+ reverse_prob=0.4,
+ noise_prob=0.4,
+ jitter_prob=0.4,
+ blur_prob=0.4,
+ hsv_aug_prob=0.4,
+ **kwargs):
+ self.crop_prob = crop_prob
+ self.reverse_prob = reverse_prob
+ self.noise_prob = noise_prob
+ self.jitter_prob = jitter_prob
+ self.blur_prob = blur_prob
+ self.hsv_aug_prob = hsv_aug_prob
+
+ def __call__(self, data):
+ img = data['image']
+ h, w, _ = img.shape
+
+ if random.random() <= self.crop_prob and h >= 20 and w >= 20:
+ img = get_crop(img)
+
+ if random.random() <= self.blur_prob:
+ img = blur(img)
+
+ if random.random() <= self.hsv_aug_prob:
+ img = hsv_aug(img)
+
+ if random.random() <= self.jitter_prob:
+ img = jitter(img)
+
+ if random.random() <= self.noise_prob:
+ img = add_gasuss_noise(img)
+
+ if random.random() <= self.reverse_prob:
+ img = 255 - img
+
+ data['image'] = img
+ return data
+
+
+class ABINetRecAug(object):
+ def __init__(self,
+ geometry_p=0.5,
+ deterioration_p=0.25,
+ colorjitter_p=0.25,
+ **kwargs):
+ self.transforms = Compose([
+ CVGeometry(
+ degrees=45,
+ translate=(0.0, 0.0),
+ scale=(0.5, 2.),
+ shear=(45, 15),
+ distortion=0.5,
+ p=geometry_p), CVDeterioration(
+ var=20, degrees=6, factor=4, p=deterioration_p),
+ CVColorJitter(
+ brightness=0.5,
+ contrast=0.5,
+ saturation=0.5,
+ hue=0.1,
+ p=colorjitter_p)
+ ])
+
+ def __call__(self, data):
+ img = data['image']
+ img = self.transforms(img)
data['image'] = img
return data
@@ -87,46 +180,6 @@ class ClsResizeImg(object):
return data
-class NRTRRecResizeImg(object):
- def __init__(self, image_shape, resize_type, padding=False, **kwargs):
- self.image_shape = image_shape
- self.resize_type = resize_type
- self.padding = padding
-
- def __call__(self, data):
- img = data['image']
- img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
- image_shape = self.image_shape
- if self.padding:
- imgC, imgH, imgW = image_shape
- # todo: change to 0 and modified image shape
- h = img.shape[0]
- w = img.shape[1]
- ratio = w / float(h)
- if math.ceil(imgH * ratio) > imgW:
- resized_w = imgW
- else:
- resized_w = int(math.ceil(imgH * ratio))
- resized_image = cv2.resize(img, (resized_w, imgH))
- norm_img = np.expand_dims(resized_image, -1)
- norm_img = norm_img.transpose((2, 0, 1))
- resized_image = norm_img.astype(np.float32) / 128. - 1.
- padding_im = np.zeros((imgC, imgH, imgW), dtype=np.float32)
- padding_im[:, :, 0:resized_w] = resized_image
- data['image'] = padding_im
- return data
- if self.resize_type == 'PIL':
- image_pil = Image.fromarray(np.uint8(img))
- img = image_pil.resize(self.image_shape, Image.ANTIALIAS)
- img = np.array(img)
- if self.resize_type == 'OpenCV':
- img = cv2.resize(img, self.image_shape)
- norm_img = np.expand_dims(img, -1)
- norm_img = norm_img.transpose((2, 0, 1))
- data['image'] = norm_img.astype(np.float32) / 128. - 1.
- return data
-
-
class RecResizeImg(object):
def __init__(self,
image_shape,
@@ -207,6 +260,84 @@ class PRENResizeImg(object):
return data
+class GrayRecResizeImg(object):
+ def __init__(self,
+ image_shape,
+ resize_type,
+ inter_type='Image.ANTIALIAS',
+ scale=True,
+ padding=False,
+ **kwargs):
+ self.image_shape = image_shape
+ self.resize_type = resize_type
+ self.padding = padding
+ self.inter_type = eval(inter_type)
+ self.scale = scale
+
+ def __call__(self, data):
+ img = data['image']
+ img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
+ image_shape = self.image_shape
+ if self.padding:
+ imgC, imgH, imgW = image_shape
+ # todo: change to 0 and modified image shape
+ h = img.shape[0]
+ w = img.shape[1]
+ ratio = w / float(h)
+ if math.ceil(imgH * ratio) > imgW:
+ resized_w = imgW
+ else:
+ resized_w = int(math.ceil(imgH * ratio))
+ resized_image = cv2.resize(img, (resized_w, imgH))
+ norm_img = np.expand_dims(resized_image, -1)
+ norm_img = norm_img.transpose((2, 0, 1))
+ resized_image = norm_img.astype(np.float32) / 128. - 1.
+ padding_im = np.zeros((imgC, imgH, imgW), dtype=np.float32)
+ padding_im[:, :, 0:resized_w] = resized_image
+ data['image'] = padding_im
+ return data
+ if self.resize_type == 'PIL':
+ image_pil = Image.fromarray(np.uint8(img))
+ img = image_pil.resize(self.image_shape, self.inter_type)
+ img = np.array(img)
+ if self.resize_type == 'OpenCV':
+ img = cv2.resize(img, self.image_shape)
+ norm_img = np.expand_dims(img, -1)
+ norm_img = norm_img.transpose((2, 0, 1))
+ if self.scale:
+ data['image'] = norm_img.astype(np.float32) / 128. - 1.
+ else:
+ data['image'] = norm_img.astype(np.float32) / 255.
+ return data
+
+
+class ABINetRecResizeImg(object):
+ def __init__(self, image_shape, **kwargs):
+ self.image_shape = image_shape
+
+ def __call__(self, data):
+ img = data['image']
+ norm_img, valid_ratio = resize_norm_img_abinet(img, self.image_shape)
+ data['image'] = norm_img
+ data['valid_ratio'] = valid_ratio
+ return data
+
+
+class SVTRRecResizeImg(object):
+ def __init__(self, image_shape, padding=True, **kwargs):
+ self.image_shape = image_shape
+ self.padding = padding
+
+ def __call__(self, data):
+ img = data['image']
+
+ norm_img, valid_ratio = resize_norm_img(img, self.image_shape,
+ self.padding)
+ data['image'] = norm_img
+ data['valid_ratio'] = valid_ratio
+ return data
+
+
def resize_norm_img_sar(img, image_shape, width_downsample_ratio=0.25):
imgC, imgH, imgW_min, imgW_max = image_shape
h = img.shape[0]
@@ -325,6 +456,26 @@ def resize_norm_img_srn(img, image_shape):
return np.reshape(img_black, (c, row, col)).astype(np.float32)
+def resize_norm_img_abinet(img, image_shape):
+ imgC, imgH, imgW = image_shape
+
+ resized_image = cv2.resize(
+ img, (imgW, imgH), interpolation=cv2.INTER_LINEAR)
+ resized_w = imgW
+ resized_image = resized_image.astype('float32')
+ resized_image = resized_image / 255.
+
+ mean = np.array([0.485, 0.456, 0.406])
+ std = np.array([0.229, 0.224, 0.225])
+ resized_image = (
+ resized_image - mean[None, None, ...]) / std[None, None, ...]
+ resized_image = resized_image.transpose((2, 0, 1))
+ resized_image = resized_image.astype('float32')
+
+ valid_ratio = min(1.0, float(resized_w / imgW))
+ return resized_image, valid_ratio
+
+
def srn_other_inputs(image_shape, num_heads, max_text_length):
imgC, imgH, imgW = image_shape
@@ -359,7 +510,7 @@ def flag():
return 1 if random.random() > 0.5000001 else -1
-def cvtColor(img):
+def hsv_aug(img):
"""
cvtColor
"""
@@ -427,50 +578,6 @@ def get_crop(image):
return crop_img
-class Config:
- """
- Config
- """
-
- def __init__(self, use_tia):
- self.anglex = random.random() * 30
- self.angley = random.random() * 15
- self.anglez = random.random() * 10
- self.fov = 42
- self.r = 0
- self.shearx = random.random() * 0.3
- self.sheary = random.random() * 0.05
- self.borderMode = cv2.BORDER_REPLICATE
- self.use_tia = use_tia
-
- def make(self, w, h, ang):
- """
- make
- """
- self.anglex = random.random() * 5 * flag()
- self.angley = random.random() * 5 * flag()
- self.anglez = -1 * random.random() * int(ang) * flag()
- self.fov = 42
- self.r = 0
- self.shearx = 0
- self.sheary = 0
- self.borderMode = cv2.BORDER_REPLICATE
- self.w = w
- self.h = h
-
- self.perspective = self.use_tia
- self.stretch = self.use_tia
- self.distort = self.use_tia
-
- self.crop = True
- self.affine = False
- self.reverse = True
- self.noise = True
- self.jitter = True
- self.blur = True
- self.color = True
-
-
def rad(x):
"""
rad
@@ -554,48 +661,3 @@ def get_warpAffine(config):
rz = np.array([[np.cos(rad(anglez)), np.sin(rad(anglez)), 0],
[-np.sin(rad(anglez)), np.cos(rad(anglez)), 0]], np.float32)
return rz
-
-
-def warp(img, ang, use_tia=True, prob=0.4):
- """
- warp
- """
- h, w, _ = img.shape
- config = Config(use_tia=use_tia)
- config.make(w, h, ang)
- new_img = img
-
- if config.distort:
- img_height, img_width = img.shape[0:2]
- if random.random() <= prob and img_height >= 20 and img_width >= 20:
- new_img = tia_distort(new_img, random.randint(3, 6))
-
- if config.stretch:
- img_height, img_width = img.shape[0:2]
- if random.random() <= prob and img_height >= 20 and img_width >= 20:
- new_img = tia_stretch(new_img, random.randint(3, 6))
-
- if config.perspective:
- if random.random() <= prob:
- new_img = tia_perspective(new_img)
-
- if config.crop:
- img_height, img_width = img.shape[0:2]
- if random.random() <= prob and img_height >= 20 and img_width >= 20:
- new_img = get_crop(new_img)
-
- if config.blur:
- if random.random() <= prob:
- new_img = blur(new_img)
- if config.color:
- if random.random() <= prob:
- new_img = cvtColor(new_img)
- if config.jitter:
- new_img = jitter(new_img)
- if config.noise:
- if random.random() <= prob:
- new_img = add_gasuss_noise(new_img)
- if config.reverse:
- if random.random() <= prob:
- new_img = 255 - new_img
- return new_img
diff --git a/ppocr/data/simple_dataset.py b/ppocr/data/simple_dataset.py
index b5da9b8898..402f1e38fe 100644
--- a/ppocr/data/simple_dataset.py
+++ b/ppocr/data/simple_dataset.py
@@ -33,7 +33,7 @@ class SimpleDataSet(Dataset):
self.delimiter = dataset_config.get('delimiter', '\t')
label_file_list = dataset_config.pop('label_file_list')
data_source_num = len(label_file_list)
- ratio_list = dataset_config.get("ratio_list", [1.0])
+ ratio_list = dataset_config.get("ratio_list", 1.0)
if isinstance(ratio_list, (float, int)):
ratio_list = [float(ratio_list)] * int(data_source_num)
diff --git a/ppocr/losses/__init__.py b/ppocr/losses/__init__.py
index de8419b7c1..7bea87f62f 100755
--- a/ppocr/losses/__init__.py
+++ b/ppocr/losses/__init__.py
@@ -30,7 +30,7 @@ from .det_fce_loss import FCELoss
from .rec_ctc_loss import CTCLoss
from .rec_att_loss import AttentionLoss
from .rec_srn_loss import SRNLoss
-from .rec_nrtr_loss import NRTRLoss
+from .rec_ce_loss import CELoss
from .rec_sar_loss import SARLoss
from .rec_aster_loss import AsterLoss
from .rec_pren_loss import PRENLoss
@@ -60,7 +60,7 @@ def build_loss(config):
support_dict = [
'DBLoss', 'PSELoss', 'EASTLoss', 'SASTLoss', 'FCELoss', 'CTCLoss',
'ClsLoss', 'AttentionLoss', 'SRNLoss', 'PGLoss', 'CombinedLoss',
- 'NRTRLoss', 'TableAttentionLoss', 'SARLoss', 'AsterLoss', 'SDMGRLoss',
+ 'CELoss', 'TableAttentionLoss', 'SARLoss', 'AsterLoss', 'SDMGRLoss',
'VQASerTokenLayoutLMLoss', 'LossFromOutput', 'PRENLoss', 'MultiLoss'
]
config = copy.deepcopy(config)
diff --git a/ppocr/losses/rec_aster_loss.py b/ppocr/losses/rec_aster_loss.py
index fbb99d29a6..52605e46db 100644
--- a/ppocr/losses/rec_aster_loss.py
+++ b/ppocr/losses/rec_aster_loss.py
@@ -27,12 +27,12 @@ class CosineEmbeddingLoss(nn.Layer):
self.epsilon = 1e-12
def forward(self, x1, x2, target):
- similarity = paddle.fluid.layers.reduce_sum(
+ similarity = paddle.sum(
x1 * x2, dim=-1) / (paddle.norm(
x1, axis=-1) * paddle.norm(
x2, axis=-1) + self.epsilon)
one_list = paddle.full_like(target, fill_value=1)
- out = paddle.fluid.layers.reduce_mean(
+ out = paddle.mean(
paddle.where(
paddle.equal(target, one_list), 1. - similarity,
paddle.maximum(
diff --git a/ppocr/losses/rec_ce_loss.py b/ppocr/losses/rec_ce_loss.py
new file mode 100644
index 0000000000..614384de86
--- /dev/null
+++ b/ppocr/losses/rec_ce_loss.py
@@ -0,0 +1,66 @@
+import paddle
+from paddle import nn
+import paddle.nn.functional as F
+
+
+class CELoss(nn.Layer):
+ def __init__(self,
+ smoothing=False,
+ with_all=False,
+ ignore_index=-1,
+ **kwargs):
+ super(CELoss, self).__init__()
+ if ignore_index >= 0:
+ self.loss_func = nn.CrossEntropyLoss(
+ reduction='mean', ignore_index=ignore_index)
+ else:
+ self.loss_func = nn.CrossEntropyLoss(reduction='mean')
+ self.smoothing = smoothing
+ self.with_all = with_all
+
+ def forward(self, pred, batch):
+
+ if isinstance(pred, dict): # for ABINet
+ loss = {}
+ loss_sum = []
+ for name, logits in pred.items():
+ if isinstance(logits, list):
+ logit_num = len(logits)
+ all_tgt = paddle.concat([batch[1]] * logit_num, 0)
+ all_logits = paddle.concat(logits, 0)
+ flt_logtis = all_logits.reshape([-1, all_logits.shape[2]])
+ flt_tgt = all_tgt.reshape([-1])
+ else:
+ flt_logtis = logits.reshape([-1, logits.shape[2]])
+ flt_tgt = batch[1].reshape([-1])
+ loss[name + '_loss'] = self.loss_func(flt_logtis, flt_tgt)
+ loss_sum.append(loss[name + '_loss'])
+ loss['loss'] = sum(loss_sum)
+ return loss
+ else:
+ if self.with_all: # for ViTSTR
+ tgt = batch[1]
+ pred = pred.reshape([-1, pred.shape[2]])
+ tgt = tgt.reshape([-1])
+ loss = self.loss_func(pred, tgt)
+ return {'loss': loss}
+ else: # for NRTR
+ max_len = batch[2].max()
+ tgt = batch[1][:, 1:2 + max_len]
+ pred = pred.reshape([-1, pred.shape[2]])
+ tgt = tgt.reshape([-1])
+ if self.smoothing:
+ eps = 0.1
+ n_class = pred.shape[1]
+ one_hot = F.one_hot(tgt, pred.shape[1])
+ one_hot = one_hot * (1 - eps) + (1 - one_hot) * eps / (
+ n_class - 1)
+ log_prb = F.log_softmax(pred, axis=1)
+ non_pad_mask = paddle.not_equal(
+ tgt, paddle.zeros(
+ tgt.shape, dtype=tgt.dtype))
+ loss = -(one_hot * log_prb).sum(axis=1)
+ loss = loss.masked_select(non_pad_mask).mean()
+ else:
+ loss = self.loss_func(pred, tgt)
+ return {'loss': loss}
diff --git a/ppocr/losses/rec_nrtr_loss.py b/ppocr/losses/rec_nrtr_loss.py
deleted file mode 100644
index 200a6d0486..0000000000
--- a/ppocr/losses/rec_nrtr_loss.py
+++ /dev/null
@@ -1,30 +0,0 @@
-import paddle
-from paddle import nn
-import paddle.nn.functional as F
-
-
-class NRTRLoss(nn.Layer):
- def __init__(self, smoothing=True, **kwargs):
- super(NRTRLoss, self).__init__()
- self.loss_func = nn.CrossEntropyLoss(reduction='mean', ignore_index=0)
- self.smoothing = smoothing
-
- def forward(self, pred, batch):
- pred = pred.reshape([-1, pred.shape[2]])
- max_len = batch[2].max()
- tgt = batch[1][:, 1:2 + max_len]
- tgt = tgt.reshape([-1])
- if self.smoothing:
- eps = 0.1
- n_class = pred.shape[1]
- one_hot = F.one_hot(tgt, pred.shape[1])
- one_hot = one_hot * (1 - eps) + (1 - one_hot) * eps / (n_class - 1)
- log_prb = F.log_softmax(pred, axis=1)
- non_pad_mask = paddle.not_equal(
- tgt, paddle.zeros(
- tgt.shape, dtype=tgt.dtype))
- loss = -(one_hot * log_prb).sum(axis=1)
- loss = loss.masked_select(non_pad_mask).mean()
- else:
- loss = self.loss_func(pred, tgt)
- return {'loss': loss}
diff --git a/ppocr/losses/table_att_loss.py b/ppocr/losses/table_att_loss.py
index d7fd99e695..51377efa2b 100644
--- a/ppocr/losses/table_att_loss.py
+++ b/ppocr/losses/table_att_loss.py
@@ -19,7 +19,6 @@ from __future__ import print_function
import paddle
from paddle import nn
from paddle.nn import functional as F
-from paddle import fluid
class TableAttentionLoss(nn.Layer):
def __init__(self, structure_weight, loc_weight, use_giou=False, giou_weight=1.0, **kwargs):
@@ -36,13 +35,13 @@ class TableAttentionLoss(nn.Layer):
:param bbox:[[x1,y1,x2,y2], [x1,y1,x2,y2],,,]
:return: loss
'''
- ix1 = fluid.layers.elementwise_max(preds[:, 0], bbox[:, 0])
- iy1 = fluid.layers.elementwise_max(preds[:, 1], bbox[:, 1])
- ix2 = fluid.layers.elementwise_min(preds[:, 2], bbox[:, 2])
- iy2 = fluid.layers.elementwise_min(preds[:, 3], bbox[:, 3])
+ ix1 = paddle.maximum(preds[:, 0], bbox[:, 0])
+ iy1 = paddle.maximum(preds[:, 1], bbox[:, 1])
+ ix2 = paddle.minimum(preds[:, 2], bbox[:, 2])
+ iy2 = paddle.minimum(preds[:, 3], bbox[:, 3])
- iw = fluid.layers.clip(ix2 - ix1 + 1e-3, 0., 1e10)
- ih = fluid.layers.clip(iy2 - iy1 + 1e-3, 0., 1e10)
+ iw = paddle.clip(ix2 - ix1 + 1e-3, 0., 1e10)
+ ih = paddle.clip(iy2 - iy1 + 1e-3, 0., 1e10)
# overlap
inters = iw * ih
@@ -55,12 +54,12 @@ class TableAttentionLoss(nn.Layer):
# ious
ious = inters / uni
- ex1 = fluid.layers.elementwise_min(preds[:, 0], bbox[:, 0])
- ey1 = fluid.layers.elementwise_min(preds[:, 1], bbox[:, 1])
- ex2 = fluid.layers.elementwise_max(preds[:, 2], bbox[:, 2])
- ey2 = fluid.layers.elementwise_max(preds[:, 3], bbox[:, 3])
- ew = fluid.layers.clip(ex2 - ex1 + 1e-3, 0., 1e10)
- eh = fluid.layers.clip(ey2 - ey1 + 1e-3, 0., 1e10)
+ ex1 = paddle.minimum(preds[:, 0], bbox[:, 0])
+ ey1 = paddle.minimum(preds[:, 1], bbox[:, 1])
+ ex2 = paddle.maximum(preds[:, 2], bbox[:, 2])
+ ey2 = paddle.maximum(preds[:, 3], bbox[:, 3])
+ ew = paddle.clip(ex2 - ex1 + 1e-3, 0., 1e10)
+ eh = paddle.clip(ey2 - ey1 + 1e-3, 0., 1e10)
# enclose erea
enclose = ew * eh + eps
diff --git a/ppocr/modeling/architectures/__init__.py b/ppocr/modeling/architectures/__init__.py
index e9a01cf028..1c955ef3ab 100755
--- a/ppocr/modeling/architectures/__init__.py
+++ b/ppocr/modeling/architectures/__init__.py
@@ -15,10 +15,13 @@
import copy
import importlib
+from paddle.jit import to_static
+from paddle.static import InputSpec
+
from .base_model import BaseModel
from .distillation_model import DistillationModel
-__all__ = ['build_model']
+__all__ = ["build_model", "apply_to_static"]
def build_model(config):
@@ -30,3 +33,36 @@ def build_model(config):
mod = importlib.import_module(__name__)
arch = getattr(mod, name)(config)
return arch
+
+
+def apply_to_static(model, config, logger):
+ if config["Global"].get("to_static", False) is not True:
+ return model
+ assert "image_shape" in config[
+ "Global"], "image_shape must be assigned for static training mode..."
+ supported_list = ["DB", "SVTR"]
+ if config["Architecture"]["algorithm"] in ["Distillation"]:
+ algo = list(config["Architecture"]["Models"].values())[0]["algorithm"]
+ else:
+ algo = config["Architecture"]["algorithm"]
+ assert algo in supported_list, f"algorithms that supports static training must in in {supported_list} but got {algo}"
+
+ specs = [
+ InputSpec(
+ [None] + config["Global"]["image_shape"], dtype='float32')
+ ]
+
+ if algo == "SVTR":
+ specs.append([
+ InputSpec(
+ [None, config["Global"]["max_text_length"]],
+ dtype='int64'), InputSpec(
+ [None, config["Global"]["max_text_length"]], dtype='int64'),
+ InputSpec(
+ [None], dtype='int64'), InputSpec(
+ [None], dtype='float64')
+ ])
+
+ model = to_static(model, input_spec=specs)
+ logger.info("Successfully to apply @to_static with specs: {}".format(specs))
+ return model
diff --git a/ppocr/modeling/backbones/__init__.py b/ppocr/modeling/backbones/__init__.py
index 072d6e0f84..f8959e263e 100755
--- a/ppocr/modeling/backbones/__init__.py
+++ b/ppocr/modeling/backbones/__init__.py
@@ -28,35 +28,37 @@ def build_backbone(config, model_type):
from .rec_mv1_enhance import MobileNetV1Enhance
from .rec_nrtr_mtb import MTB
from .rec_resnet_31 import ResNet31
+ from .rec_resnet_45 import ResNet45
from .rec_resnet_aster import ResNet_ASTER
from .rec_micronet import MicroNet
from .rec_efficientb3_pren import EfficientNetb3_PREN
from .rec_svtrnet import SVTRNet
+ from .rec_vitstr import ViTSTR
support_dict = [
'MobileNetV1Enhance', 'MobileNetV3', 'ResNet', 'ResNetFPN', 'MTB',
- "ResNet31", "ResNet_ASTER", 'MicroNet', 'EfficientNetb3_PREN',
- 'SVTRNet'
+ 'ResNet31', 'ResNet45', 'ResNet_ASTER', 'MicroNet',
+ 'EfficientNetb3_PREN', 'SVTRNet', 'ViTSTR'
]
- elif model_type == "e2e":
+ elif model_type == 'e2e':
from .e2e_resnet_vd_pg import ResNet
support_dict = ['ResNet']
elif model_type == 'kie':
from .kie_unet_sdmgr import Kie_backbone
support_dict = ['Kie_backbone']
- elif model_type == "table":
+ elif model_type == 'table':
from .table_resnet_vd import ResNet
from .table_mobilenet_v3 import MobileNetV3
- support_dict = ["ResNet", "MobileNetV3"]
+ support_dict = ['ResNet', 'MobileNetV3']
elif model_type == 'vqa':
from .vqa_layoutlm import LayoutLMForSer, LayoutLMv2ForSer, LayoutLMv2ForRe, LayoutXLMForSer, LayoutXLMForRe
support_dict = [
- "LayoutLMForSer", "LayoutLMv2ForSer", 'LayoutLMv2ForRe',
- "LayoutXLMForSer", 'LayoutXLMForRe'
+ 'LayoutLMForSer', 'LayoutLMv2ForSer', 'LayoutLMv2ForRe',
+ 'LayoutXLMForSer', 'LayoutXLMForRe'
]
else:
raise NotImplementedError
- module_name = config.pop("name")
+ module_name = config.pop('name')
assert module_name in support_dict, Exception(
"when model typs is {}, backbone only support {}".format(model_type,
support_dict))
diff --git a/ppocr/modeling/backbones/kie_unet_sdmgr.py b/ppocr/modeling/backbones/kie_unet_sdmgr.py
index 545e4e7511..4b1bd80300 100644
--- a/ppocr/modeling/backbones/kie_unet_sdmgr.py
+++ b/ppocr/modeling/backbones/kie_unet_sdmgr.py
@@ -175,12 +175,7 @@ class Kie_backbone(nn.Layer):
img, relations, texts, gt_bboxes, tag, img_size)
x = self.img_feat(img)
boxes, rois_num = self.bbox2roi(gt_bboxes)
- feats = paddle.fluid.layers.roi_align(
- x,
- boxes,
- spatial_scale=1.0,
- pooled_height=7,
- pooled_width=7,
- rois_num=rois_num)
+ feats = paddle.vision.ops.roi_align(
+ x, boxes, spatial_scale=1.0, output_size=7, boxes_num=rois_num)
feats = self.maxpool(feats).squeeze(-1).squeeze(-1)
return [relations, texts, feats]
diff --git a/ppocr/modeling/backbones/rec_resnet_45.py b/ppocr/modeling/backbones/rec_resnet_45.py
new file mode 100644
index 0000000000..9093d0bc99
--- /dev/null
+++ b/ppocr/modeling/backbones/rec_resnet_45.py
@@ -0,0 +1,147 @@
+# copyright (c) 2021 PaddlePaddle Authors. All Rights Reserve.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+This code is refer from:
+https://github.com/FangShancheng/ABINet/tree/main/modules
+"""
+
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+
+import paddle
+from paddle import ParamAttr
+from paddle.nn.initializer import KaimingNormal
+import paddle.nn as nn
+import paddle.nn.functional as F
+import numpy as np
+import math
+
+__all__ = ["ResNet45"]
+
+
+def conv1x1(in_planes, out_planes, stride=1):
+ return nn.Conv2D(
+ in_planes,
+ out_planes,
+ kernel_size=1,
+ stride=1,
+ weight_attr=ParamAttr(initializer=KaimingNormal()),
+ bias_attr=False)
+
+
+def conv3x3(in_channel, out_channel, stride=1):
+ return nn.Conv2D(
+ in_channel,
+ out_channel,
+ kernel_size=3,
+ stride=stride,
+ padding=1,
+ weight_attr=ParamAttr(initializer=KaimingNormal()),
+ bias_attr=False)
+
+
+class BasicBlock(nn.Layer):
+ expansion = 1
+
+ def __init__(self, in_channels, channels, stride=1, downsample=None):
+ super().__init__()
+ self.conv1 = conv1x1(in_channels, channels)
+ self.bn1 = nn.BatchNorm2D(channels)
+ self.relu = nn.ReLU()
+ self.conv2 = conv3x3(channels, channels, stride)
+ self.bn2 = nn.BatchNorm2D(channels)
+ self.downsample = downsample
+ self.stride = stride
+
+ def forward(self, x):
+ residual = x
+
+ out = self.conv1(x)
+ out = self.bn1(out)
+ out = self.relu(out)
+
+ out = self.conv2(out)
+ out = self.bn2(out)
+
+ if self.downsample is not None:
+ residual = self.downsample(x)
+ out += residual
+ out = self.relu(out)
+
+ return out
+
+
+class ResNet45(nn.Layer):
+ def __init__(self, block=BasicBlock, layers=[3, 4, 6, 6, 3], in_channels=3):
+ self.inplanes = 32
+ super(ResNet45, self).__init__()
+ self.conv1 = nn.Conv2D(
+ 3,
+ 32,
+ kernel_size=3,
+ stride=1,
+ padding=1,
+ weight_attr=ParamAttr(initializer=KaimingNormal()),
+ bias_attr=False)
+ self.bn1 = nn.BatchNorm2D(32)
+ self.relu = nn.ReLU()
+
+ self.layer1 = self._make_layer(block, 32, layers[0], stride=2)
+ self.layer2 = self._make_layer(block, 64, layers[1], stride=1)
+ self.layer3 = self._make_layer(block, 128, layers[2], stride=2)
+ self.layer4 = self._make_layer(block, 256, layers[3], stride=1)
+ self.layer5 = self._make_layer(block, 512, layers[4], stride=1)
+ self.out_channels = 512
+
+ # for m in self.modules():
+ # if isinstance(m, nn.Conv2D):
+ # n = m._kernel_size[0] * m._kernel_size[1] * m._out_channels
+ # m.weight.data.normal_(0, math.sqrt(2. / n))
+
+ def _make_layer(self, block, planes, blocks, stride=1):
+ downsample = None
+ if stride != 1 or self.inplanes != planes * block.expansion:
+ # downsample = True
+ downsample = nn.Sequential(
+ nn.Conv2D(
+ self.inplanes,
+ planes * block.expansion,
+ kernel_size=1,
+ stride=stride,
+ weight_attr=ParamAttr(initializer=KaimingNormal()),
+ bias_attr=False),
+ nn.BatchNorm2D(planes * block.expansion), )
+
+ layers = []
+ layers.append(block(self.inplanes, planes, stride, downsample))
+ self.inplanes = planes * block.expansion
+ for i in range(1, blocks):
+ layers.append(block(self.inplanes, planes))
+
+ return nn.Sequential(*layers)
+
+ def forward(self, x):
+
+ x = self.conv1(x)
+ x = self.bn1(x)
+ x = self.relu(x)
+ # print(x)
+ x = self.layer1(x)
+ x = self.layer2(x)
+ x = self.layer3(x)
+ # print(x)
+ x = self.layer4(x)
+ x = self.layer5(x)
+ return x
diff --git a/ppocr/modeling/backbones/rec_resnet_fpn.py b/ppocr/modeling/backbones/rec_resnet_fpn.py
index a7e876a2bd..79efd6e41e 100644
--- a/ppocr/modeling/backbones/rec_resnet_fpn.py
+++ b/ppocr/modeling/backbones/rec_resnet_fpn.py
@@ -18,7 +18,6 @@ from __future__ import print_function
from paddle import nn, ParamAttr
from paddle.nn import functional as F
-import paddle.fluid as fluid
import paddle
import numpy as np
diff --git a/ppocr/modeling/backbones/rec_svtrnet.py b/ppocr/modeling/backbones/rec_svtrnet.py
index c57bf46345..c2c07f4476 100644
--- a/ppocr/modeling/backbones/rec_svtrnet.py
+++ b/ppocr/modeling/backbones/rec_svtrnet.py
@@ -147,7 +147,7 @@ class Attention(nn.Layer):
dim,
num_heads=8,
mixer='Global',
- HW=[8, 25],
+ HW=None,
local_k=[7, 11],
qkv_bias=False,
qk_scale=None,
@@ -210,7 +210,7 @@ class Block(nn.Layer):
num_heads,
mixer='Global',
local_mixer=[7, 11],
- HW=[8, 25],
+ HW=None,
mlp_ratio=4.,
qkv_bias=False,
qk_scale=None,
@@ -274,7 +274,9 @@ class PatchEmbed(nn.Layer):
img_size=[32, 100],
in_channels=3,
embed_dim=768,
- sub_num=2):
+ sub_num=2,
+ patch_size=[4, 4],
+ mode='pope'):
super().__init__()
num_patches = (img_size[1] // (2 ** sub_num)) * \
(img_size[0] // (2 ** sub_num))
@@ -282,50 +284,56 @@ class PatchEmbed(nn.Layer):
self.num_patches = num_patches
self.embed_dim = embed_dim
self.norm = None
- if sub_num == 2:
- self.proj = nn.Sequential(
- ConvBNLayer(
- in_channels=in_channels,
- out_channels=embed_dim // 2,
- kernel_size=3,
- stride=2,
- padding=1,
- act=nn.GELU,
- bias_attr=None),
- ConvBNLayer(
- in_channels=embed_dim // 2,
- out_channels=embed_dim,
- kernel_size=3,
- stride=2,
- padding=1,
- act=nn.GELU,
- bias_attr=None))
- if sub_num == 3:
- self.proj = nn.Sequential(
- ConvBNLayer(
- in_channels=in_channels,
- out_channels=embed_dim // 4,
- kernel_size=3,
- stride=2,
- padding=1,
- act=nn.GELU,
- bias_attr=None),
- ConvBNLayer(
- in_channels=embed_dim // 4,
- out_channels=embed_dim // 2,
- kernel_size=3,
- stride=2,
- padding=1,
- act=nn.GELU,
- bias_attr=None),
- ConvBNLayer(
- in_channels=embed_dim // 2,
- out_channels=embed_dim,
- kernel_size=3,
- stride=2,
- padding=1,
- act=nn.GELU,
- bias_attr=None))
+ if mode == 'pope':
+ if sub_num == 2:
+ self.proj = nn.Sequential(
+ ConvBNLayer(
+ in_channels=in_channels,
+ out_channels=embed_dim // 2,
+ kernel_size=3,
+ stride=2,
+ padding=1,
+ act=nn.GELU,
+ bias_attr=None),
+ ConvBNLayer(
+ in_channels=embed_dim // 2,
+ out_channels=embed_dim,
+ kernel_size=3,
+ stride=2,
+ padding=1,
+ act=nn.GELU,
+ bias_attr=None))
+ if sub_num == 3:
+ self.proj = nn.Sequential(
+ ConvBNLayer(
+ in_channels=in_channels,
+ out_channels=embed_dim // 4,
+ kernel_size=3,
+ stride=2,
+ padding=1,
+ act=nn.GELU,
+ bias_attr=None),
+ ConvBNLayer(
+ in_channels=embed_dim // 4,
+ out_channels=embed_dim // 2,
+ kernel_size=3,
+ stride=2,
+ padding=1,
+ act=nn.GELU,
+ bias_attr=None),
+ ConvBNLayer(
+ in_channels=embed_dim // 2,
+ out_channels=embed_dim,
+ kernel_size=3,
+ stride=2,
+ padding=1,
+ act=nn.GELU,
+ bias_attr=None))
+ elif mode == 'linear':
+ self.proj = nn.Conv2D(
+ 1, embed_dim, kernel_size=patch_size, stride=patch_size)
+ self.num_patches = img_size[0] // patch_size[0] * img_size[
+ 1] // patch_size[1]
def forward(self, x):
B, C, H, W = x.shape
diff --git a/ppocr/modeling/backbones/rec_vitstr.py b/ppocr/modeling/backbones/rec_vitstr.py
new file mode 100644
index 0000000000..d5d7d5148a
--- /dev/null
+++ b/ppocr/modeling/backbones/rec_vitstr.py
@@ -0,0 +1,120 @@
+# copyright (c) 2021 PaddlePaddle Authors. All Rights Reserve.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+This code is refer from:
+https://github.com/roatienza/deep-text-recognition-benchmark/blob/master/modules/vitstr.py
+"""
+
+import numpy as np
+import paddle
+import paddle.nn as nn
+from ppocr.modeling.backbones.rec_svtrnet import Block, PatchEmbed, zeros_, trunc_normal_, ones_
+
+scale_dim_heads = {'tiny': [192, 3], 'small': [384, 6], 'base': [768, 12]}
+
+
+class ViTSTR(nn.Layer):
+ def __init__(self,
+ img_size=[224, 224],
+ in_channels=1,
+ scale='tiny',
+ seqlen=27,
+ patch_size=[16, 16],
+ embed_dim=None,
+ depth=12,
+ num_heads=None,
+ mlp_ratio=4,
+ qkv_bias=True,
+ qk_scale=None,
+ drop_path_rate=0.,
+ drop_rate=0.,
+ attn_drop_rate=0.,
+ norm_layer='nn.LayerNorm',
+ act_layer='nn.GELU',
+ epsilon=1e-6,
+ out_channels=None,
+ **kwargs):
+ super().__init__()
+ self.seqlen = seqlen
+ embed_dim = embed_dim if embed_dim is not None else scale_dim_heads[
+ scale][0]
+ num_heads = num_heads if num_heads is not None else scale_dim_heads[
+ scale][1]
+ out_channels = out_channels if out_channels is not None else embed_dim
+ self.patch_embed = PatchEmbed(
+ img_size=img_size,
+ in_channels=in_channels,
+ embed_dim=embed_dim,
+ patch_size=patch_size,
+ mode='linear')
+ num_patches = self.patch_embed.num_patches
+
+ self.pos_embed = self.create_parameter(
+ shape=[1, num_patches + 1, embed_dim], default_initializer=zeros_)
+ self.add_parameter("pos_embed", self.pos_embed)
+ self.cls_token = self.create_parameter(
+ shape=[1, 1, embed_dim], default_initializer=zeros_)
+ self.add_parameter("cls_token", self.cls_token)
+
+ self.pos_drop = nn.Dropout(p=drop_rate)
+
+ dpr = np.linspace(0, drop_path_rate, depth)
+ self.blocks = nn.LayerList([
+ Block(
+ dim=embed_dim,
+ num_heads=num_heads,
+ mlp_ratio=mlp_ratio,
+ qkv_bias=qkv_bias,
+ qk_scale=qk_scale,
+ drop=drop_rate,
+ attn_drop=attn_drop_rate,
+ drop_path=dpr[i],
+ norm_layer=norm_layer,
+ act_layer=eval(act_layer),
+ epsilon=epsilon,
+ prenorm=False) for i in range(depth)
+ ])
+ self.norm = eval(norm_layer)(embed_dim, epsilon=epsilon)
+
+ self.out_channels = out_channels
+
+ trunc_normal_(self.pos_embed)
+ trunc_normal_(self.cls_token)
+ self.apply(self._init_weights)
+
+ def _init_weights(self, m):
+ if isinstance(m, nn.Linear):
+ trunc_normal_(m.weight)
+ if isinstance(m, nn.Linear) and m.bias is not None:
+ zeros_(m.bias)
+ elif isinstance(m, nn.LayerNorm):
+ zeros_(m.bias)
+ ones_(m.weight)
+
+ def forward_features(self, x):
+ B = x.shape[0]
+ x = self.patch_embed(x)
+ cls_tokens = paddle.tile(self.cls_token, repeat_times=[B, 1, 1])
+ x = paddle.concat((cls_tokens, x), axis=1)
+ x = x + self.pos_embed
+ x = self.pos_drop(x)
+ for blk in self.blocks:
+ x = blk(x)
+ x = self.norm(x)
+ return x
+
+ def forward(self, x):
+ x = self.forward_features(x)
+ x = x[:, :self.seqlen]
+ return x.transpose([0, 2, 1]).unsqueeze(2)
diff --git a/ppocr/modeling/heads/__init__.py b/ppocr/modeling/heads/__init__.py
index 1670ea38e6..14e6aab854 100755
--- a/ppocr/modeling/heads/__init__.py
+++ b/ppocr/modeling/heads/__init__.py
@@ -33,6 +33,7 @@ def build_head(config):
from .rec_aster_head import AsterHead
from .rec_pren_head import PRENHead
from .rec_multi_head import MultiHead
+ from .rec_abinet_head import ABINetHead
# cls head
from .cls_head import ClsHead
@@ -46,7 +47,7 @@ def build_head(config):
'DBHead', 'PSEHead', 'FCEHead', 'EASTHead', 'SASTHead', 'CTCHead',
'ClsHead', 'AttentionHead', 'SRNHead', 'PGHead', 'Transformer',
'TableAttentionHead', 'SARHead', 'AsterHead', 'SDMGRHead', 'PRENHead',
- 'MultiHead'
+ 'MultiHead', 'ABINetHead'
]
#table head
diff --git a/ppocr/modeling/heads/multiheadAttention.py b/ppocr/modeling/heads/multiheadAttention.py
deleted file mode 100755
index 900865ba1a..0000000000
--- a/ppocr/modeling/heads/multiheadAttention.py
+++ /dev/null
@@ -1,163 +0,0 @@
-# copyright (c) 2021 PaddlePaddle Authors. All Rights Reserve.
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-# http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-
-import paddle
-from paddle import nn
-import paddle.nn.functional as F
-from paddle.nn import Linear
-from paddle.nn.initializer import XavierUniform as xavier_uniform_
-from paddle.nn.initializer import Constant as constant_
-from paddle.nn.initializer import XavierNormal as xavier_normal_
-
-zeros_ = constant_(value=0.)
-ones_ = constant_(value=1.)
-
-
-class MultiheadAttention(nn.Layer):
- """Allows the model to jointly attend to information
- from different representation subspaces.
- See reference: Attention Is All You Need
-
- .. math::
- \text{MultiHead}(Q, K, V) = \text{Concat}(head_1,\dots,head_h)W^O
- \text{where} head_i = \text{Attention}(QW_i^Q, KW_i^K, VW_i^V)
-
- Args:
- embed_dim: total dimension of the model
- num_heads: parallel attention layers, or heads
-
- """
-
- def __init__(self,
- embed_dim,
- num_heads,
- dropout=0.,
- bias=True,
- add_bias_kv=False,
- add_zero_attn=False):
- super(MultiheadAttention, self).__init__()
- self.embed_dim = embed_dim
- self.num_heads = num_heads
- self.dropout = dropout
- self.head_dim = embed_dim // num_heads
- assert self.head_dim * num_heads == self.embed_dim, "embed_dim must be divisible by num_heads"
- self.scaling = self.head_dim**-0.5
- self.out_proj = Linear(embed_dim, embed_dim, bias_attr=bias)
- self._reset_parameters()
- self.conv1 = paddle.nn.Conv2D(
- in_channels=embed_dim, out_channels=embed_dim, kernel_size=(1, 1))
- self.conv2 = paddle.nn.Conv2D(
- in_channels=embed_dim, out_channels=embed_dim, kernel_size=(1, 1))
- self.conv3 = paddle.nn.Conv2D(
- in_channels=embed_dim, out_channels=embed_dim, kernel_size=(1, 1))
-
- def _reset_parameters(self):
- xavier_uniform_(self.out_proj.weight)
-
- def forward(self,
- query,
- key,
- value,
- key_padding_mask=None,
- incremental_state=None,
- attn_mask=None):
- """
- Inputs of forward function
- query: [target length, batch size, embed dim]
- key: [sequence length, batch size, embed dim]
- value: [sequence length, batch size, embed dim]
- key_padding_mask: if True, mask padding based on batch size
- incremental_state: if provided, previous time steps are cashed
- need_weights: output attn_output_weights
- static_kv: key and value are static
-
- Outputs of forward function
- attn_output: [target length, batch size, embed dim]
- attn_output_weights: [batch size, target length, sequence length]
- """
- q_shape = paddle.shape(query)
- src_shape = paddle.shape(key)
- q = self._in_proj_q(query)
- k = self._in_proj_k(key)
- v = self._in_proj_v(value)
- q *= self.scaling
- q = paddle.transpose(
- paddle.reshape(
- q, [q_shape[0], q_shape[1], self.num_heads, self.head_dim]),
- [1, 2, 0, 3])
- k = paddle.transpose(
- paddle.reshape(
- k, [src_shape[0], q_shape[1], self.num_heads, self.head_dim]),
- [1, 2, 0, 3])
- v = paddle.transpose(
- paddle.reshape(
- v, [src_shape[0], q_shape[1], self.num_heads, self.head_dim]),
- [1, 2, 0, 3])
- if key_padding_mask is not None:
- assert key_padding_mask.shape[0] == q_shape[1]
- assert key_padding_mask.shape[1] == src_shape[0]
- attn_output_weights = paddle.matmul(q,
- paddle.transpose(k, [0, 1, 3, 2]))
- if attn_mask is not None:
- attn_mask = paddle.unsqueeze(paddle.unsqueeze(attn_mask, 0), 0)
- attn_output_weights += attn_mask
- if key_padding_mask is not None:
- attn_output_weights = paddle.reshape(
- attn_output_weights,
- [q_shape[1], self.num_heads, q_shape[0], src_shape[0]])
- key = paddle.unsqueeze(paddle.unsqueeze(key_padding_mask, 1), 2)
- key = paddle.cast(key, 'float32')
- y = paddle.full(
- shape=paddle.shape(key), dtype='float32', fill_value='-inf')
- y = paddle.where(key == 0., key, y)
- attn_output_weights += y
- attn_output_weights = F.softmax(
- attn_output_weights.astype('float32'),
- axis=-1,
- dtype=paddle.float32 if attn_output_weights.dtype == paddle.float16
- else attn_output_weights.dtype)
- attn_output_weights = F.dropout(
- attn_output_weights, p=self.dropout, training=self.training)
-
- attn_output = paddle.matmul(attn_output_weights, v)
- attn_output = paddle.reshape(
- paddle.transpose(attn_output, [2, 0, 1, 3]),
- [q_shape[0], q_shape[1], self.embed_dim])
- attn_output = self.out_proj(attn_output)
-
- return attn_output
-
- def _in_proj_q(self, query):
- query = paddle.transpose(query, [1, 2, 0])
- query = paddle.unsqueeze(query, axis=2)
- res = self.conv1(query)
- res = paddle.squeeze(res, axis=2)
- res = paddle.transpose(res, [2, 0, 1])
- return res
-
- def _in_proj_k(self, key):
- key = paddle.transpose(key, [1, 2, 0])
- key = paddle.unsqueeze(key, axis=2)
- res = self.conv2(key)
- res = paddle.squeeze(res, axis=2)
- res = paddle.transpose(res, [2, 0, 1])
- return res
-
- def _in_proj_v(self, value):
- value = paddle.transpose(value, [1, 2, 0]) #(1, 2, 0)
- value = paddle.unsqueeze(value, axis=2)
- res = self.conv3(value)
- res = paddle.squeeze(res, axis=2)
- res = paddle.transpose(res, [2, 0, 1])
- return res
diff --git a/ppocr/modeling/heads/rec_abinet_head.py b/ppocr/modeling/heads/rec_abinet_head.py
new file mode 100644
index 0000000000..a0f60f1be1
--- /dev/null
+++ b/ppocr/modeling/heads/rec_abinet_head.py
@@ -0,0 +1,296 @@
+# copyright (c) 2021 PaddlePaddle Authors. All Rights Reserve.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+This code is refer from:
+https://github.com/FangShancheng/ABINet/tree/main/modules
+"""
+
+import math
+import paddle
+from paddle import nn
+import paddle.nn.functional as F
+from paddle.nn import LayerList
+from ppocr.modeling.heads.rec_nrtr_head import TransformerBlock, PositionalEncoding
+
+
+class BCNLanguage(nn.Layer):
+ def __init__(self,
+ d_model=512,
+ nhead=8,
+ num_layers=4,
+ dim_feedforward=2048,
+ dropout=0.,
+ max_length=25,
+ detach=True,
+ num_classes=37):
+ super().__init__()
+
+ self.d_model = d_model
+ self.detach = detach
+ self.max_length = max_length + 1 # additional stop token
+ self.proj = nn.Linear(num_classes, d_model, bias_attr=False)
+ self.token_encoder = PositionalEncoding(
+ dropout=0.1, dim=d_model, max_len=self.max_length)
+ self.pos_encoder = PositionalEncoding(
+ dropout=0, dim=d_model, max_len=self.max_length)
+
+ self.decoder = nn.LayerList([
+ TransformerBlock(
+ d_model=d_model,
+ nhead=nhead,
+ dim_feedforward=dim_feedforward,
+ attention_dropout_rate=dropout,
+ residual_dropout_rate=dropout,
+ with_self_attn=False,
+ with_cross_attn=True) for i in range(num_layers)
+ ])
+
+ self.cls = nn.Linear(d_model, num_classes)
+
+ def forward(self, tokens, lengths):
+ """
+ Args:
+ tokens: (B, N, C) where N is length, B is batch size and C is classes number
+ lengths: (B,)
+ """
+ if self.detach: tokens = tokens.detach()
+ embed = self.proj(tokens) # (B, N, C)
+ embed = self.token_encoder(embed) # (B, N, C)
+ padding_mask = _get_mask(lengths, self.max_length)
+ zeros = paddle.zeros_like(embed) # (B, N, C)
+ qeury = self.pos_encoder(zeros)
+ for decoder_layer in self.decoder:
+ qeury = decoder_layer(qeury, embed, cross_mask=padding_mask)
+ output = qeury # (B, N, C)
+
+ logits = self.cls(output) # (B, N, C)
+
+ return output, logits
+
+
+def encoder_layer(in_c, out_c, k=3, s=2, p=1):
+ return nn.Sequential(
+ nn.Conv2D(in_c, out_c, k, s, p), nn.BatchNorm2D(out_c), nn.ReLU())
+
+
+def decoder_layer(in_c,
+ out_c,
+ k=3,
+ s=1,
+ p=1,
+ mode='nearest',
+ scale_factor=None,
+ size=None):
+ align_corners = False if mode == 'nearest' else True
+ return nn.Sequential(
+ nn.Upsample(
+ size=size,
+ scale_factor=scale_factor,
+ mode=mode,
+ align_corners=align_corners),
+ nn.Conv2D(in_c, out_c, k, s, p),
+ nn.BatchNorm2D(out_c),
+ nn.ReLU())
+
+
+class PositionAttention(nn.Layer):
+ def __init__(self,
+ max_length,
+ in_channels=512,
+ num_channels=64,
+ h=8,
+ w=32,
+ mode='nearest',
+ **kwargs):
+ super().__init__()
+ self.max_length = max_length
+ self.k_encoder = nn.Sequential(
+ encoder_layer(
+ in_channels, num_channels, s=(1, 2)),
+ encoder_layer(
+ num_channels, num_channels, s=(2, 2)),
+ encoder_layer(
+ num_channels, num_channels, s=(2, 2)),
+ encoder_layer(
+ num_channels, num_channels, s=(2, 2)))
+ self.k_decoder = nn.Sequential(
+ decoder_layer(
+ num_channels, num_channels, scale_factor=2, mode=mode),
+ decoder_layer(
+ num_channels, num_channels, scale_factor=2, mode=mode),
+ decoder_layer(
+ num_channels, num_channels, scale_factor=2, mode=mode),
+ decoder_layer(
+ num_channels, in_channels, size=(h, w), mode=mode))
+
+ self.pos_encoder = PositionalEncoding(
+ dropout=0, dim=in_channels, max_len=max_length)
+ self.project = nn.Linear(in_channels, in_channels)
+
+ def forward(self, x):
+ B, C, H, W = x.shape
+ k, v = x, x
+
+ # calculate key vector
+ features = []
+ for i in range(0, len(self.k_encoder)):
+ k = self.k_encoder[i](k)
+ features.append(k)
+ for i in range(0, len(self.k_decoder) - 1):
+ k = self.k_decoder[i](k)
+ # print(k.shape, features[len(self.k_decoder) - 2 - i].shape)
+ k = k + features[len(self.k_decoder) - 2 - i]
+ k = self.k_decoder[-1](k)
+
+ # calculate query vector
+ # TODO q=f(q,k)
+ zeros = paddle.zeros(
+ (B, self.max_length, C), dtype=x.dtype) # (T, N, C)
+ q = self.pos_encoder(zeros) # (B, N, C)
+ q = self.project(q) # (B, N, C)
+
+ # calculate attention
+ attn_scores = q @k.flatten(2) # (B, N, (H*W))
+ attn_scores = attn_scores / (C**0.5)
+ attn_scores = F.softmax(attn_scores, axis=-1)
+
+ v = v.flatten(2).transpose([0, 2, 1]) # (B, (H*W), C)
+ attn_vecs = attn_scores @v # (B, N, C)
+
+ return attn_vecs, attn_scores.reshape([0, self.max_length, H, W])
+
+
+class ABINetHead(nn.Layer):
+ def __init__(self,
+ in_channels,
+ out_channels,
+ d_model=512,
+ nhead=8,
+ num_layers=3,
+ dim_feedforward=2048,
+ dropout=0.1,
+ max_length=25,
+ use_lang=False,
+ iter_size=1):
+ super().__init__()
+ self.max_length = max_length + 1
+ self.pos_encoder = PositionalEncoding(
+ dropout=0.1, dim=d_model, max_len=8 * 32)
+ self.encoder = nn.LayerList([
+ TransformerBlock(
+ d_model=d_model,
+ nhead=nhead,
+ dim_feedforward=dim_feedforward,
+ attention_dropout_rate=dropout,
+ residual_dropout_rate=dropout,
+ with_self_attn=True,
+ with_cross_attn=False) for i in range(num_layers)
+ ])
+ self.decoder = PositionAttention(
+ max_length=max_length + 1, # additional stop token
+ mode='nearest', )
+ self.out_channels = out_channels
+ self.cls = nn.Linear(d_model, self.out_channels)
+ self.use_lang = use_lang
+ if use_lang:
+ self.iter_size = iter_size
+ self.language = BCNLanguage(
+ d_model=d_model,
+ nhead=nhead,
+ num_layers=4,
+ dim_feedforward=dim_feedforward,
+ dropout=dropout,
+ max_length=max_length,
+ num_classes=self.out_channels)
+ # alignment
+ self.w_att_align = nn.Linear(2 * d_model, d_model)
+ self.cls_align = nn.Linear(d_model, self.out_channels)
+
+ def forward(self, x, targets=None):
+ x = x.transpose([0, 2, 3, 1])
+ _, H, W, C = x.shape
+ feature = x.flatten(1, 2)
+ feature = self.pos_encoder(feature)
+ for encoder_layer in self.encoder:
+ feature = encoder_layer(feature)
+ feature = feature.reshape([0, H, W, C]).transpose([0, 3, 1, 2])
+ v_feature, attn_scores = self.decoder(
+ feature) # (B, N, C), (B, C, H, W)
+ vis_logits = self.cls(v_feature) # (B, N, C)
+ logits = vis_logits
+ vis_lengths = _get_length(vis_logits)
+ if self.use_lang:
+ align_logits = vis_logits
+ align_lengths = vis_lengths
+ all_l_res, all_a_res = [], []
+ for i in range(self.iter_size):
+ tokens = F.softmax(align_logits, axis=-1)
+ lengths = align_lengths
+ lengths = paddle.clip(
+ lengths, 2, self.max_length) # TODO:move to langauge model
+ l_feature, l_logits = self.language(tokens, lengths)
+
+ # alignment
+ all_l_res.append(l_logits)
+ fuse = paddle.concat((l_feature, v_feature), -1)
+ f_att = F.sigmoid(self.w_att_align(fuse))
+ output = f_att * v_feature + (1 - f_att) * l_feature
+ align_logits = self.cls_align(output) # (B, N, C)
+
+ align_lengths = _get_length(align_logits)
+ all_a_res.append(align_logits)
+ if self.training:
+ return {
+ 'align': all_a_res,
+ 'lang': all_l_res,
+ 'vision': vis_logits
+ }
+ else:
+ logits = align_logits
+ if self.training:
+ return logits
+ else:
+ return F.softmax(logits, -1)
+
+
+def _get_length(logit):
+ """ Greed decoder to obtain length from logit"""
+ out = (logit.argmax(-1) == 0)
+ abn = out.any(-1)
+ out_int = out.cast('int32')
+ out = (out_int.cumsum(-1) == 1) & out
+ out = out.cast('int32')
+ out = out.argmax(-1)
+ out = out + 1
+ out = paddle.where(abn, out, paddle.to_tensor(logit.shape[1]))
+ return out
+
+
+def _get_mask(length, max_length):
+ """Generate a square mask for the sequence. The masked positions are filled with float('-inf').
+ Unmasked positions are filled with float(0.0).
+ """
+ length = length.unsqueeze(-1)
+ B = paddle.shape(length)[0]
+ grid = paddle.arange(0, max_length).unsqueeze(0).tile([B, 1])
+ zero_mask = paddle.zeros([B, max_length], dtype='float32')
+ inf_mask = paddle.full([B, max_length], '-inf', dtype='float32')
+ diag_mask = paddle.diag(
+ paddle.full(
+ [max_length], '-inf', dtype=paddle.float32),
+ offset=0,
+ name=None)
+ mask = paddle.where(grid >= length, inf_mask, zero_mask)
+ mask = mask.unsqueeze(1) + diag_mask
+ return mask.unsqueeze(1)
diff --git a/ppocr/modeling/heads/rec_nrtr_head.py b/ppocr/modeling/heads/rec_nrtr_head.py
index 38ba0c9178..bf9ef56145 100644
--- a/ppocr/modeling/heads/rec_nrtr_head.py
+++ b/ppocr/modeling/heads/rec_nrtr_head.py
@@ -14,20 +14,15 @@
import math
import paddle
-import copy
from paddle import nn
import paddle.nn.functional as F
from paddle.nn import LayerList
-from paddle.nn.initializer import XavierNormal as xavier_uniform_
-from paddle.nn import Dropout, Linear, LayerNorm, Conv2D
+# from paddle.nn.initializer import XavierNormal as xavier_uniform_
+from paddle.nn import Dropout, Linear, LayerNorm
import numpy as np
-from ppocr.modeling.heads.multiheadAttention import MultiheadAttention
-from paddle.nn.initializer import Constant as constant_
+from ppocr.modeling.backbones.rec_svtrnet import Mlp, zeros_, ones_
from paddle.nn.initializer import XavierNormal as xavier_normal_
-zeros_ = constant_(value=0.)
-ones_ = constant_(value=1.)
-
class Transformer(nn.Layer):
"""A transformer model. User is able to modify the attributes as needed. The architechture
@@ -45,7 +40,6 @@ class Transformer(nn.Layer):
dropout: the dropout value (default=0.1).
custom_encoder: custom encoder (default=None).
custom_decoder: custom decoder (default=None).
-
"""
def __init__(self,
@@ -54,45 +48,49 @@ class Transformer(nn.Layer):
num_encoder_layers=6,
beam_size=0,
num_decoder_layers=6,
+ max_len=25,
dim_feedforward=1024,
attention_dropout_rate=0.0,
residual_dropout_rate=0.1,
- custom_encoder=None,
- custom_decoder=None,
in_channels=0,
out_channels=0,
scale_embedding=True):
super(Transformer, self).__init__()
self.out_channels = out_channels + 1
+ self.max_len = max_len
self.embedding = Embeddings(
d_model=d_model,
vocab=self.out_channels,
padding_idx=0,
scale_embedding=scale_embedding)
self.positional_encoding = PositionalEncoding(
- dropout=residual_dropout_rate,
- dim=d_model, )
- if custom_encoder is not None:
- self.encoder = custom_encoder
- else:
- if num_encoder_layers > 0:
- encoder_layer = TransformerEncoderLayer(
- d_model, nhead, dim_feedforward, attention_dropout_rate,
- residual_dropout_rate)
- self.encoder = TransformerEncoder(encoder_layer,
- num_encoder_layers)
- else:
- self.encoder = None
+ dropout=residual_dropout_rate, dim=d_model)
- if custom_decoder is not None:
- self.decoder = custom_decoder
+ if num_encoder_layers > 0:
+ self.encoder = nn.LayerList([
+ TransformerBlock(
+ d_model,
+ nhead,
+ dim_feedforward,
+ attention_dropout_rate,
+ residual_dropout_rate,
+ with_self_attn=True,
+ with_cross_attn=False) for i in range(num_encoder_layers)
+ ])
else:
- decoder_layer = TransformerDecoderLayer(
- d_model, nhead, dim_feedforward, attention_dropout_rate,
- residual_dropout_rate)
- self.decoder = TransformerDecoder(decoder_layer, num_decoder_layers)
+ self.encoder = None
+
+ self.decoder = nn.LayerList([
+ TransformerBlock(
+ d_model,
+ nhead,
+ dim_feedforward,
+ attention_dropout_rate,
+ residual_dropout_rate,
+ with_self_attn=True,
+ with_cross_attn=True) for i in range(num_decoder_layers)
+ ])
- self._reset_parameters()
self.beam_size = beam_size
self.d_model = d_model
self.nhead = nhead
@@ -105,7 +103,7 @@ class Transformer(nn.Layer):
def _init_weights(self, m):
- if isinstance(m, nn.Conv2D):
+ if isinstance(m, nn.Linear):
xavier_normal_(m.weight)
if m.bias is not None:
zeros_(m.bias)
@@ -113,24 +111,20 @@ class Transformer(nn.Layer):
def forward_train(self, src, tgt):
tgt = tgt[:, :-1]
- tgt_key_padding_mask = self.generate_padding_mask(tgt)
- tgt = self.embedding(tgt).transpose([1, 0, 2])
+ tgt = self.embedding(tgt)
tgt = self.positional_encoding(tgt)
- tgt_mask = self.generate_square_subsequent_mask(tgt.shape[0])
+ tgt_mask = self.generate_square_subsequent_mask(tgt.shape[1])
if self.encoder is not None:
- src = self.positional_encoding(src.transpose([1, 0, 2]))
- memory = self.encoder(src)
+ src = self.positional_encoding(src)
+ for encoder_layer in self.encoder:
+ src = encoder_layer(src)
+ memory = src # B N C
else:
- memory = src.squeeze(2).transpose([2, 0, 1])
- output = self.decoder(
- tgt,
- memory,
- tgt_mask=tgt_mask,
- memory_mask=None,
- tgt_key_padding_mask=tgt_key_padding_mask,
- memory_key_padding_mask=None)
- output = output.transpose([1, 0, 2])
+ memory = src # B N C
+ for decoder_layer in self.decoder:
+ tgt = decoder_layer(tgt, memory, self_mask=tgt_mask)
+ output = tgt
logit = self.tgt_word_prj(output)
return logit
@@ -140,8 +134,8 @@ class Transformer(nn.Layer):
src: the sequence to the encoder (required).
tgt: the sequence to the decoder (required).
Shape:
- - src: :math:`(S, N, E)`.
- - tgt: :math:`(T, N, E)`.
+ - src: :math:`(B, sN, C)`.
+ - tgt: :math:`(B, tN, C)`.
Examples:
>>> output = transformer_model(src, tgt)
"""
@@ -157,36 +151,35 @@ class Transformer(nn.Layer):
return self.forward_test(src)
def forward_test(self, src):
+
bs = paddle.shape(src)[0]
if self.encoder is not None:
- src = self.positional_encoding(paddle.transpose(src, [1, 0, 2]))
- memory = self.encoder(src)
+ src = self.positional_encoding(src)
+ for encoder_layer in self.encoder:
+ src = encoder_layer(src)
+ memory = src # B N C
else:
- memory = paddle.transpose(paddle.squeeze(src, 2), [2, 0, 1])
+ memory = src
dec_seq = paddle.full((bs, 1), 2, dtype=paddle.int64)
dec_prob = paddle.full((bs, 1), 1., dtype=paddle.float32)
- for len_dec_seq in range(1, 25):
- dec_seq_embed = paddle.transpose(self.embedding(dec_seq), [1, 0, 2])
+ for len_dec_seq in range(1, self.max_len):
+ dec_seq_embed = self.embedding(dec_seq)
dec_seq_embed = self.positional_encoding(dec_seq_embed)
tgt_mask = self.generate_square_subsequent_mask(
- paddle.shape(dec_seq_embed)[0])
- output = self.decoder(
- dec_seq_embed,
- memory,
- tgt_mask=tgt_mask,
- memory_mask=None,
- tgt_key_padding_mask=None,
- memory_key_padding_mask=None)
- dec_output = paddle.transpose(output, [1, 0, 2])
+ paddle.shape(dec_seq_embed)[1])
+ tgt = dec_seq_embed
+ for decoder_layer in self.decoder:
+ tgt = decoder_layer(tgt, memory, self_mask=tgt_mask)
+ dec_output = tgt
dec_output = dec_output[:, -1, :]
- word_prob = F.softmax(self.tgt_word_prj(dec_output), axis=1)
- preds_idx = paddle.argmax(word_prob, axis=1)
+ word_prob = F.softmax(self.tgt_word_prj(dec_output), axis=-1)
+ preds_idx = paddle.argmax(word_prob, axis=-1)
if paddle.equal_all(
preds_idx,
paddle.full(
paddle.shape(preds_idx), 3, dtype='int64')):
break
- preds_prob = paddle.max(word_prob, axis=1)
+ preds_prob = paddle.max(word_prob, axis=-1)
dec_seq = paddle.concat(
[dec_seq, paddle.reshape(preds_idx, [-1, 1])], axis=1)
dec_prob = paddle.concat(
@@ -194,10 +187,10 @@ class Transformer(nn.Layer):
return [dec_seq, dec_prob]
def forward_beam(self, images):
- ''' Translation work in one batch '''
+ """ Translation work in one batch """
def get_inst_idx_to_tensor_position_map(inst_idx_list):
- ''' Indicate the position of an instance in a tensor. '''
+ """ Indicate the position of an instance in a tensor. """
return {
inst_idx: tensor_position
for tensor_position, inst_idx in enumerate(inst_idx_list)
@@ -205,7 +198,7 @@ class Transformer(nn.Layer):
def collect_active_part(beamed_tensor, curr_active_inst_idx,
n_prev_active_inst, n_bm):
- ''' Collect tensor parts associated to active instances. '''
+ """ Collect tensor parts associated to active instances. """
beamed_tensor_shape = paddle.shape(beamed_tensor)
n_curr_active_inst = len(curr_active_inst_idx)
@@ -237,9 +230,8 @@ class Transformer(nn.Layer):
return active_src_enc, active_inst_idx_to_position_map
def beam_decode_step(inst_dec_beams, len_dec_seq, enc_output,
- inst_idx_to_position_map, n_bm,
- memory_key_padding_mask):
- ''' Decode and update beam status, and then return active beam idx '''
+ inst_idx_to_position_map, n_bm):
+ """ Decode and update beam status, and then return active beam idx """
def prepare_beam_dec_seq(inst_dec_beams, len_dec_seq):
dec_partial_seq = [
@@ -249,19 +241,15 @@ class Transformer(nn.Layer):
dec_partial_seq = dec_partial_seq.reshape([-1, len_dec_seq])
return dec_partial_seq
- def predict_word(dec_seq, enc_output, n_active_inst, n_bm,
- memory_key_padding_mask):
- dec_seq = paddle.transpose(self.embedding(dec_seq), [1, 0, 2])
+ def predict_word(dec_seq, enc_output, n_active_inst, n_bm):
+ dec_seq = self.embedding(dec_seq)
dec_seq = self.positional_encoding(dec_seq)
tgt_mask = self.generate_square_subsequent_mask(
- paddle.shape(dec_seq)[0])
- dec_output = self.decoder(
- dec_seq,
- enc_output,
- tgt_mask=tgt_mask,
- tgt_key_padding_mask=None,
- memory_key_padding_mask=memory_key_padding_mask, )
- dec_output = paddle.transpose(dec_output, [1, 0, 2])
+ paddle.shape(dec_seq)[1])
+ tgt = dec_seq
+ for decoder_layer in self.decoder:
+ tgt = decoder_layer(tgt, enc_output, self_mask=tgt_mask)
+ dec_output = tgt
dec_output = dec_output[:,
-1, :] # Pick the last step: (bh * bm) * d_h
word_prob = F.softmax(self.tgt_word_prj(dec_output), axis=1)
@@ -281,8 +269,7 @@ class Transformer(nn.Layer):
n_active_inst = len(inst_idx_to_position_map)
dec_seq = prepare_beam_dec_seq(inst_dec_beams, len_dec_seq)
- word_prob = predict_word(dec_seq, enc_output, n_active_inst, n_bm,
- None)
+ word_prob = predict_word(dec_seq, enc_output, n_active_inst, n_bm)
# Update the beam with predicted word prob information and collect incomplete instances
active_inst_idx_list = collect_active_inst_idx_list(
inst_dec_beams, word_prob, inst_idx_to_position_map)
@@ -303,10 +290,10 @@ class Transformer(nn.Layer):
with paddle.no_grad():
#-- Encode
if self.encoder is not None:
- src = self.positional_encoding(images.transpose([1, 0, 2]))
+ src = self.positional_encoding(images)
src_enc = self.encoder(src)
else:
- src_enc = images.squeeze(2).transpose([0, 2, 1])
+ src_enc = images
n_bm = self.beam_size
src_shape = paddle.shape(src_enc)
@@ -317,11 +304,11 @@ class Transformer(nn.Layer):
inst_idx_to_position_map = get_inst_idx_to_tensor_position_map(
active_inst_idx_list)
# Decode
- for len_dec_seq in range(1, 25):
+ for len_dec_seq in range(1, self.max_len):
src_enc_copy = src_enc.clone()
active_inst_idx_list = beam_decode_step(
inst_dec_beams, len_dec_seq, src_enc_copy,
- inst_idx_to_position_map, n_bm, None)
+ inst_idx_to_position_map, n_bm)
if not active_inst_idx_list:
break # all instances have finished their path to
src_enc, inst_idx_to_position_map = collate_active_info(
@@ -354,261 +341,124 @@ class Transformer(nn.Layer):
shape=[sz, sz], dtype='float32', fill_value='-inf'),
diagonal=1)
mask = mask + mask_inf
- return mask
-
- def generate_padding_mask(self, x):
- padding_mask = paddle.equal(x, paddle.to_tensor(0, dtype=x.dtype))
- return padding_mask
-
- def _reset_parameters(self):
- """Initiate parameters in the transformer model."""
-
- for p in self.parameters():
- if p.dim() > 1:
- xavier_uniform_(p)
+ return mask.unsqueeze([0, 1])
-class TransformerEncoder(nn.Layer):
- """TransformerEncoder is a stack of N encoder layers
- Args:
- encoder_layer: an instance of the TransformerEncoderLayer() class (required).
- num_layers: the number of sub-encoder-layers in the encoder (required).
- norm: the layer normalization component (optional).
- """
+class MultiheadAttention(nn.Layer):
+ """Allows the model to jointly attend to information
+ from different representation subspaces.
+ See reference: Attention Is All You Need
- def __init__(self, encoder_layer, num_layers):
- super(TransformerEncoder, self).__init__()
- self.layers = _get_clones(encoder_layer, num_layers)
- self.num_layers = num_layers
-
- def forward(self, src):
- """Pass the input through the endocder layers in turn.
- Args:
- src: the sequnce to the encoder (required).
- mask: the mask for the src sequence (optional).
- src_key_padding_mask: the mask for the src keys per batch (optional).
- """
- output = src
-
- for i in range(self.num_layers):
- output = self.layers[i](output,
- src_mask=None,
- src_key_padding_mask=None)
-
- return output
-
-
-class TransformerDecoder(nn.Layer):
- """TransformerDecoder is a stack of N decoder layers
+ .. math::
+ \text{MultiHead}(Q, K, V) = \text{Concat}(head_1,\dots,head_h)W^O
+ \text{where} head_i = \text{Attention}(QW_i^Q, KW_i^K, VW_i^V)
Args:
- decoder_layer: an instance of the TransformerDecoderLayer() class (required).
- num_layers: the number of sub-decoder-layers in the decoder (required).
- norm: the layer normalization component (optional).
+ embed_dim: total dimension of the model
+ num_heads: parallel attention layers, or heads
"""
- def __init__(self, decoder_layer, num_layers):
- super(TransformerDecoder, self).__init__()
- self.layers = _get_clones(decoder_layer, num_layers)
- self.num_layers = num_layers
+ def __init__(self, embed_dim, num_heads, dropout=0., self_attn=False):
+ super(MultiheadAttention, self).__init__()
+ self.embed_dim = embed_dim
+ self.num_heads = num_heads
+ # self.dropout = dropout
+ self.head_dim = embed_dim // num_heads
+ assert self.head_dim * num_heads == self.embed_dim, "embed_dim must be divisible by num_heads"
+ self.scale = self.head_dim**-0.5
+ self.self_attn = self_attn
+ if self_attn:
+ self.qkv = nn.Linear(embed_dim, embed_dim * 3)
+ else:
+ self.q = nn.Linear(embed_dim, embed_dim)
+ self.kv = nn.Linear(embed_dim, embed_dim * 2)
+ self.attn_drop = nn.Dropout(dropout)
+ self.out_proj = nn.Linear(embed_dim, embed_dim)
- def forward(self,
- tgt,
- memory,
- tgt_mask=None,
- memory_mask=None,
- tgt_key_padding_mask=None,
- memory_key_padding_mask=None):
- """Pass the inputs (and mask) through the decoder layer in turn.
+ def forward(self, query, key=None, attn_mask=None):
- Args:
- tgt: the sequence to the decoder (required).
- memory: the sequnce from the last layer of the encoder (required).
- tgt_mask: the mask for the tgt sequence (optional).
- memory_mask: the mask for the memory sequence (optional).
- tgt_key_padding_mask: the mask for the tgt keys per batch (optional).
- memory_key_padding_mask: the mask for the memory keys per batch (optional).
- """
- output = tgt
- for i in range(self.num_layers):
- output = self.layers[i](
- output,
- memory,
- tgt_mask=tgt_mask,
- memory_mask=memory_mask,
- tgt_key_padding_mask=tgt_key_padding_mask,
- memory_key_padding_mask=memory_key_padding_mask)
+ qN = query.shape[1]
- return output
+ if self.self_attn:
+ qkv = self.qkv(query).reshape(
+ (0, qN, 3, self.num_heads, self.head_dim)).transpose(
+ (2, 0, 3, 1, 4))
+ q, k, v = qkv[0], qkv[1], qkv[2]
+ else:
+ kN = key.shape[1]
+ q = self.q(query).reshape(
+ [0, qN, self.num_heads, self.head_dim]).transpose([0, 2, 1, 3])
+ kv = self.kv(key).reshape(
+ (0, kN, 2, self.num_heads, self.head_dim)).transpose(
+ (2, 0, 3, 1, 4))
+ k, v = kv[0], kv[1]
+
+ attn = (q.matmul(k.transpose((0, 1, 3, 2)))) * self.scale
+
+ if attn_mask is not None:
+ attn += attn_mask
+
+ attn = F.softmax(attn, axis=-1)
+ attn = self.attn_drop(attn)
+
+ x = (attn.matmul(v)).transpose((0, 2, 1, 3)).reshape(
+ (0, qN, self.embed_dim))
+ x = self.out_proj(x)
+
+ return x
-class TransformerEncoderLayer(nn.Layer):
- """TransformerEncoderLayer is made up of self-attn and feedforward network.
- This standard encoder layer is based on the paper "Attention Is All You Need".
- Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez,
- Lukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. In Advances in
- Neural Information Processing Systems, pages 6000-6010. Users may modify or implement
- in a different way during application.
-
- Args:
- d_model: the number of expected features in the input (required).
- nhead: the number of heads in the multiheadattention models (required).
- dim_feedforward: the dimension of the feedforward network model (default=2048).
- dropout: the dropout value (default=0.1).
-
- """
-
+class TransformerBlock(nn.Layer):
def __init__(self,
d_model,
nhead,
dim_feedforward=2048,
attention_dropout_rate=0.0,
- residual_dropout_rate=0.1):
- super(TransformerEncoderLayer, self).__init__()
- self.self_attn = MultiheadAttention(
- d_model, nhead, dropout=attention_dropout_rate)
+ residual_dropout_rate=0.1,
+ with_self_attn=True,
+ with_cross_attn=False,
+ epsilon=1e-5):
+ super(TransformerBlock, self).__init__()
+ self.with_self_attn = with_self_attn
+ if with_self_attn:
+ self.self_attn = MultiheadAttention(
+ d_model,
+ nhead,
+ dropout=attention_dropout_rate,
+ self_attn=with_self_attn)
+ self.norm1 = LayerNorm(d_model, epsilon=epsilon)
+ self.dropout1 = Dropout(residual_dropout_rate)
+ self.with_cross_attn = with_cross_attn
+ if with_cross_attn:
+ self.cross_attn = MultiheadAttention( #for self_attn of encoder or cross_attn of decoder
+ d_model,
+ nhead,
+ dropout=attention_dropout_rate)
+ self.norm2 = LayerNorm(d_model, epsilon=epsilon)
+ self.dropout2 = Dropout(residual_dropout_rate)
- self.conv1 = Conv2D(
- in_channels=d_model,
- out_channels=dim_feedforward,
- kernel_size=(1, 1))
- self.conv2 = Conv2D(
- in_channels=dim_feedforward,
- out_channels=d_model,
- kernel_size=(1, 1))
+ self.mlp = Mlp(in_features=d_model,
+ hidden_features=dim_feedforward,
+ act_layer=nn.ReLU,
+ drop=residual_dropout_rate)
- self.norm1 = LayerNorm(d_model)
- self.norm2 = LayerNorm(d_model)
- self.dropout1 = Dropout(residual_dropout_rate)
- self.dropout2 = Dropout(residual_dropout_rate)
+ self.norm3 = LayerNorm(d_model, epsilon=epsilon)
- def forward(self, src, src_mask=None, src_key_padding_mask=None):
- """Pass the input through the endocder layer.
- Args:
- src: the sequnce to the encoder layer (required).
- src_mask: the mask for the src sequence (optional).
- src_key_padding_mask: the mask for the src keys per batch (optional).
- """
- src2 = self.self_attn(
- src,
- src,
- src,
- attn_mask=src_mask,
- key_padding_mask=src_key_padding_mask)
- src = src + self.dropout1(src2)
- src = self.norm1(src)
-
- src = paddle.transpose(src, [1, 2, 0])
- src = paddle.unsqueeze(src, 2)
- src2 = self.conv2(F.relu(self.conv1(src)))
- src2 = paddle.squeeze(src2, 2)
- src2 = paddle.transpose(src2, [2, 0, 1])
- src = paddle.squeeze(src, 2)
- src = paddle.transpose(src, [2, 0, 1])
-
- src = src + self.dropout2(src2)
- src = self.norm2(src)
- return src
-
-
-class TransformerDecoderLayer(nn.Layer):
- """TransformerDecoderLayer is made up of self-attn, multi-head-attn and feedforward network.
- This standard decoder layer is based on the paper "Attention Is All You Need".
- Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez,
- Lukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. In Advances in
- Neural Information Processing Systems, pages 6000-6010. Users may modify or implement
- in a different way during application.
-
- Args:
- d_model: the number of expected features in the input (required).
- nhead: the number of heads in the multiheadattention models (required).
- dim_feedforward: the dimension of the feedforward network model (default=2048).
- dropout: the dropout value (default=0.1).
-
- """
-
- def __init__(self,
- d_model,
- nhead,
- dim_feedforward=2048,
- attention_dropout_rate=0.0,
- residual_dropout_rate=0.1):
- super(TransformerDecoderLayer, self).__init__()
- self.self_attn = MultiheadAttention(
- d_model, nhead, dropout=attention_dropout_rate)
- self.multihead_attn = MultiheadAttention(
- d_model, nhead, dropout=attention_dropout_rate)
-
- self.conv1 = Conv2D(
- in_channels=d_model,
- out_channels=dim_feedforward,
- kernel_size=(1, 1))
- self.conv2 = Conv2D(
- in_channels=dim_feedforward,
- out_channels=d_model,
- kernel_size=(1, 1))
-
- self.norm1 = LayerNorm(d_model)
- self.norm2 = LayerNorm(d_model)
- self.norm3 = LayerNorm(d_model)
- self.dropout1 = Dropout(residual_dropout_rate)
- self.dropout2 = Dropout(residual_dropout_rate)
self.dropout3 = Dropout(residual_dropout_rate)
- def forward(self,
- tgt,
- memory,
- tgt_mask=None,
- memory_mask=None,
- tgt_key_padding_mask=None,
- memory_key_padding_mask=None):
- """Pass the inputs (and mask) through the decoder layer.
+ def forward(self, tgt, memory=None, self_mask=None, cross_mask=None):
+ if self.with_self_attn:
+ tgt1 = self.self_attn(tgt, attn_mask=self_mask)
+ tgt = self.norm1(tgt + self.dropout1(tgt1))
- Args:
- tgt: the sequence to the decoder layer (required).
- memory: the sequnce from the last layer of the encoder (required).
- tgt_mask: the mask for the tgt sequence (optional).
- memory_mask: the mask for the memory sequence (optional).
- tgt_key_padding_mask: the mask for the tgt keys per batch (optional).
- memory_key_padding_mask: the mask for the memory keys per batch (optional).
-
- """
- tgt2 = self.self_attn(
- tgt,
- tgt,
- tgt,
- attn_mask=tgt_mask,
- key_padding_mask=tgt_key_padding_mask)
- tgt = tgt + self.dropout1(tgt2)
- tgt = self.norm1(tgt)
- tgt2 = self.multihead_attn(
- tgt,
- memory,
- memory,
- attn_mask=memory_mask,
- key_padding_mask=memory_key_padding_mask)
- tgt = tgt + self.dropout2(tgt2)
- tgt = self.norm2(tgt)
-
- # default
- tgt = paddle.transpose(tgt, [1, 2, 0])
- tgt = paddle.unsqueeze(tgt, 2)
- tgt2 = self.conv2(F.relu(self.conv1(tgt)))
- tgt2 = paddle.squeeze(tgt2, 2)
- tgt2 = paddle.transpose(tgt2, [2, 0, 1])
- tgt = paddle.squeeze(tgt, 2)
- tgt = paddle.transpose(tgt, [2, 0, 1])
-
- tgt = tgt + self.dropout3(tgt2)
- tgt = self.norm3(tgt)
+ if self.with_cross_attn:
+ tgt2 = self.cross_attn(tgt, key=memory, attn_mask=cross_mask)
+ tgt = self.norm2(tgt + self.dropout2(tgt2))
+ tgt = self.norm3(tgt + self.dropout3(self.mlp(tgt)))
return tgt
-def _get_clones(module, N):
- return LayerList([copy.deepcopy(module) for i in range(N)])
-
-
class PositionalEncoding(nn.Layer):
"""Inject some information about the relative or absolute position of the tokens
in the sequence. The positional encodings have the same dimension as
@@ -651,8 +501,9 @@ class PositionalEncoding(nn.Layer):
Examples:
>>> output = pos_encoder(x)
"""
+ x = x.transpose([1, 0, 2])
x = x + self.pe[:paddle.shape(x)[0], :]
- return self.dropout(x)
+ return self.dropout(x).transpose([1, 0, 2])
class PositionalEncoding_2d(nn.Layer):
@@ -725,7 +576,7 @@ class PositionalEncoding_2d(nn.Layer):
class Embeddings(nn.Layer):
- def __init__(self, d_model, vocab, padding_idx, scale_embedding):
+ def __init__(self, d_model, vocab, padding_idx=None, scale_embedding=True):
super(Embeddings, self).__init__()
self.embedding = nn.Embedding(vocab, d_model, padding_idx=padding_idx)
w0 = np.random.normal(0.0, d_model**-0.5,
@@ -742,7 +593,7 @@ class Embeddings(nn.Layer):
class Beam():
- ''' Beam search '''
+ """ Beam search """
def __init__(self, size, device=False):
diff --git a/ppocr/modeling/heads/rec_sar_head.py b/ppocr/modeling/heads/rec_sar_head.py
index 0e6b34404b..5e64cae85a 100644
--- a/ppocr/modeling/heads/rec_sar_head.py
+++ b/ppocr/modeling/heads/rec_sar_head.py
@@ -83,7 +83,7 @@ class SAREncoder(nn.Layer):
def forward(self, feat, img_metas=None):
if img_metas is not None:
- assert len(img_metas[0]) == feat.shape[0]
+ assert len(img_metas[0]) == paddle.shape(feat)[0]
valid_ratios = None
if img_metas is not None and self.mask:
@@ -98,9 +98,10 @@ class SAREncoder(nn.Layer):
if valid_ratios is not None:
valid_hf = []
- T = holistic_feat.shape[1]
- for i in range(len(valid_ratios)):
- valid_step = min(T, math.ceil(T * valid_ratios[i])) - 1
+ T = paddle.shape(holistic_feat)[1]
+ for i in range(paddle.shape(valid_ratios)[0]):
+ valid_step = paddle.minimum(
+ T, paddle.ceil(valid_ratios[i] * T).astype('int32')) - 1
valid_hf.append(holistic_feat[i, valid_step, :])
valid_hf = paddle.stack(valid_hf, axis=0)
else:
@@ -247,13 +248,14 @@ class ParallelSARDecoder(BaseDecoder):
# bsz * (seq_len + 1) * h * w * attn_size
attn_weight = self.conv1x1_2(attn_weight)
# bsz * (seq_len + 1) * h * w * 1
- bsz, T, h, w, c = attn_weight.shape
+ bsz, T, h, w, c = paddle.shape(attn_weight)
assert c == 1
if valid_ratios is not None:
# cal mask of attention weight
- for i in range(len(valid_ratios)):
- valid_width = min(w, math.ceil(w * valid_ratios[i]))
+ for i in range(paddle.shape(valid_ratios)[0]):
+ valid_width = paddle.minimum(
+ w, paddle.ceil(valid_ratios[i] * w).astype("int32"))
if valid_width < w:
attn_weight[i, :, :, valid_width:, :] = float('-inf')
@@ -288,7 +290,7 @@ class ParallelSARDecoder(BaseDecoder):
img_metas: [label, valid_ratio]
'''
if img_metas is not None:
- assert len(img_metas[0]) == feat.shape[0]
+ assert paddle.shape(img_metas[0])[0] == paddle.shape(feat)[0]
valid_ratios = None
if img_metas is not None and self.mask:
@@ -302,7 +304,6 @@ class ParallelSARDecoder(BaseDecoder):
# bsz * (seq_len + 1) * C
out_dec = self._2d_attention(
in_dec, feat, out_enc, valid_ratios=valid_ratios)
- # bsz * (seq_len + 1) * num_classes
return out_dec[:, 1:, :] # bsz * seq_len * num_classes
@@ -395,7 +396,6 @@ class SARHead(nn.Layer):
if self.training:
label = targets[0] # label
- label = paddle.to_tensor(label, dtype='int64')
final_out = self.decoder(
feat, holistic_feat, label, img_metas=targets)
else:
diff --git a/ppocr/modeling/heads/rec_srn_head.py b/ppocr/modeling/heads/rec_srn_head.py
index 8d59e4711a..1070d8cd64 100644
--- a/ppocr/modeling/heads/rec_srn_head.py
+++ b/ppocr/modeling/heads/rec_srn_head.py
@@ -20,13 +20,11 @@ import math
import paddle
from paddle import nn, ParamAttr
from paddle.nn import functional as F
-import paddle.fluid as fluid
import numpy as np
from .self_attention import WrapEncoderForFeature
from .self_attention import WrapEncoder
from paddle.static import Program
from ppocr.modeling.backbones.rec_resnet_fpn import ResNetFPN
-import paddle.fluid.framework as framework
from collections import OrderedDict
gradient_clip = 10
diff --git a/ppocr/modeling/heads/self_attention.py b/ppocr/modeling/heads/self_attention.py
index 6c27fdbe43..6e4c65e393 100644
--- a/ppocr/modeling/heads/self_attention.py
+++ b/ppocr/modeling/heads/self_attention.py
@@ -22,7 +22,6 @@ import paddle
from paddle import ParamAttr, nn
from paddle import nn, ParamAttr
from paddle.nn import functional as F
-import paddle.fluid as fluid
import numpy as np
gradient_clip = 10
@@ -288,10 +287,10 @@ class PrePostProcessLayer(nn.Layer):
"layer_norm_%d" % len(self.sublayers()),
paddle.nn.LayerNorm(
normalized_shape=d_model,
- weight_attr=fluid.ParamAttr(
- initializer=fluid.initializer.Constant(1.)),
- bias_attr=fluid.ParamAttr(
- initializer=fluid.initializer.Constant(0.)))))
+ weight_attr=paddle.ParamAttr(
+ initializer=paddle.nn.initializer.Constant(1.)),
+ bias_attr=paddle.ParamAttr(
+ initializer=paddle.nn.initializer.Constant(0.)))))
elif cmd == "d": # add dropout
self.functors.append(lambda x: F.dropout(
x, p=dropout_rate, mode="downscale_in_infer")
@@ -324,7 +323,7 @@ class PrepareEncoder(nn.Layer):
def forward(self, src_word, src_pos):
src_word_emb = src_word
- src_word_emb = fluid.layers.cast(src_word_emb, 'float32')
+ src_word_emb = paddle.cast(src_word_emb, 'float32')
src_word_emb = paddle.scale(x=src_word_emb, scale=self.src_emb_dim**0.5)
src_pos = paddle.squeeze(src_pos, axis=-1)
src_pos_enc = self.emb(src_pos)
@@ -367,7 +366,7 @@ class PrepareDecoder(nn.Layer):
self.dropout_rate = dropout_rate
def forward(self, src_word, src_pos):
- src_word = fluid.layers.cast(src_word, 'int64')
+ src_word = paddle.cast(src_word, 'int64')
src_word = paddle.squeeze(src_word, axis=-1)
src_word_emb = self.emb0(src_word)
src_word_emb = paddle.scale(x=src_word_emb, scale=self.src_emb_dim**0.5)
diff --git a/ppocr/postprocess/__init__.py b/ppocr/postprocess/__init__.py
index f50b5f1c5f..2635117c84 100644
--- a/ppocr/postprocess/__init__.py
+++ b/ppocr/postprocess/__init__.py
@@ -27,7 +27,7 @@ from .sast_postprocess import SASTPostProcess
from .fce_postprocess import FCEPostProcess
from .rec_postprocess import CTCLabelDecode, AttnLabelDecode, SRNLabelDecode, \
DistillationCTCLabelDecode, TableLabelDecode, NRTRLabelDecode, SARLabelDecode, \
- SEEDLabelDecode, PRENLabelDecode
+ SEEDLabelDecode, PRENLabelDecode, ViTSTRLabelDecode, ABINetLabelDecode
from .cls_postprocess import ClsPostProcess
from .pg_postprocess import PGPostProcess
from .vqa_token_ser_layoutlm_postprocess import VQASerTokenLayoutLMPostProcess
@@ -42,7 +42,7 @@ def build_post_process(config, global_config=None):
'DistillationDBPostProcess', 'NRTRLabelDecode', 'SARLabelDecode',
'SEEDLabelDecode', 'VQASerTokenLayoutLMPostProcess',
'VQAReTokenLayoutLMPostProcess', 'PRENLabelDecode',
- 'DistillationSARLabelDecode'
+ 'DistillationSARLabelDecode', 'ViTSTRLabelDecode', 'ABINetLabelDecode'
]
if config['name'] == 'PSEPostProcess':
diff --git a/ppocr/postprocess/rec_postprocess.py b/ppocr/postprocess/rec_postprocess.py
index bf0fd890bf..c77420ad19 100644
--- a/ppocr/postprocess/rec_postprocess.py
+++ b/ppocr/postprocess/rec_postprocess.py
@@ -140,70 +140,6 @@ class DistillationCTCLabelDecode(CTCLabelDecode):
return output
-class NRTRLabelDecode(BaseRecLabelDecode):
- """ Convert between text-label and text-index """
-
- def __init__(self, character_dict_path=None, use_space_char=True, **kwargs):
- super(NRTRLabelDecode, self).__init__(character_dict_path,
- use_space_char)
-
- def __call__(self, preds, label=None, *args, **kwargs):
-
- if len(preds) == 2:
- preds_id = preds[0]
- preds_prob = preds[1]
- if isinstance(preds_id, paddle.Tensor):
- preds_id = preds_id.numpy()
- if isinstance(preds_prob, paddle.Tensor):
- preds_prob = preds_prob.numpy()
- if preds_id[0][0] == 2:
- preds_idx = preds_id[:, 1:]
- preds_prob = preds_prob[:, 1:]
- else:
- preds_idx = preds_id
- text = self.decode(preds_idx, preds_prob, is_remove_duplicate=False)
- if label is None:
- return text
- label = self.decode(label[:, 1:])
- else:
- if isinstance(preds, paddle.Tensor):
- preds = preds.numpy()
- preds_idx = preds.argmax(axis=2)
- preds_prob = preds.max(axis=2)
- text = self.decode(preds_idx, preds_prob, is_remove_duplicate=False)
- if label is None:
- return text
- label = self.decode(label[:, 1:])
- return text, label
-
- def add_special_char(self, dict_character):
- dict_character = ['blank', '', '', ''] + dict_character
- return dict_character
-
- def decode(self, text_index, text_prob=None, is_remove_duplicate=False):
- """ convert text-index into text-label. """
- result_list = []
- batch_size = len(text_index)
- for batch_idx in range(batch_size):
- char_list = []
- conf_list = []
- for idx in range(len(text_index[batch_idx])):
- if text_index[batch_idx][idx] == 3: # end
- break
- try:
- char_list.append(self.character[int(text_index[batch_idx][
- idx])])
- except:
- continue
- if text_prob is not None:
- conf_list.append(text_prob[batch_idx][idx])
- else:
- conf_list.append(1)
- text = ''.join(char_list)
- result_list.append((text.lower(), np.mean(conf_list).tolist()))
- return result_list
-
-
class AttnLabelDecode(BaseRecLabelDecode):
""" Convert between text-label and text-index """
@@ -752,3 +688,122 @@ class PRENLabelDecode(BaseRecLabelDecode):
return text
label = self.decode(label)
return text, label
+
+
+class NRTRLabelDecode(BaseRecLabelDecode):
+ """ Convert between text-label and text-index """
+
+ def __init__(self, character_dict_path=None, use_space_char=True, **kwargs):
+ super(NRTRLabelDecode, self).__init__(character_dict_path,
+ use_space_char)
+
+ def __call__(self, preds, label=None, *args, **kwargs):
+
+ if len(preds) == 2:
+ preds_id = preds[0]
+ preds_prob = preds[1]
+ if isinstance(preds_id, paddle.Tensor):
+ preds_id = preds_id.numpy()
+ if isinstance(preds_prob, paddle.Tensor):
+ preds_prob = preds_prob.numpy()
+ if preds_id[0][0] == 2:
+ preds_idx = preds_id[:, 1:]
+ preds_prob = preds_prob[:, 1:]
+ else:
+ preds_idx = preds_id
+ text = self.decode(preds_idx, preds_prob, is_remove_duplicate=False)
+ if label is None:
+ return text
+ label = self.decode(label[:, 1:])
+ else:
+ if isinstance(preds, paddle.Tensor):
+ preds = preds.numpy()
+ preds_idx = preds.argmax(axis=2)
+ preds_prob = preds.max(axis=2)
+ text = self.decode(preds_idx, preds_prob, is_remove_duplicate=False)
+ if label is None:
+ return text
+ label = self.decode(label[:, 1:])
+ return text, label
+
+ def add_special_char(self, dict_character):
+ dict_character = ['blank', '', '', ''] + dict_character
+ return dict_character
+
+ def decode(self, text_index, text_prob=None, is_remove_duplicate=False):
+ """ convert text-index into text-label. """
+ result_list = []
+ batch_size = len(text_index)
+ for batch_idx in range(batch_size):
+ char_list = []
+ conf_list = []
+ for idx in range(len(text_index[batch_idx])):
+ try:
+ char_idx = self.character[int(text_index[batch_idx][idx])]
+ except:
+ continue
+ if char_idx == '': # end
+ break
+ char_list.append(char_idx)
+ if text_prob is not None:
+ conf_list.append(text_prob[batch_idx][idx])
+ else:
+ conf_list.append(1)
+ text = ''.join(char_list)
+ result_list.append((text.lower(), np.mean(conf_list).tolist()))
+ return result_list
+
+
+class ViTSTRLabelDecode(NRTRLabelDecode):
+ """ Convert between text-label and text-index """
+
+ def __init__(self, character_dict_path=None, use_space_char=False,
+ **kwargs):
+ super(ViTSTRLabelDecode, self).__init__(character_dict_path,
+ use_space_char)
+
+ def __call__(self, preds, label=None, *args, **kwargs):
+ if isinstance(preds, paddle.Tensor):
+ preds = preds[:, 1:].numpy()
+ else:
+ preds = preds[:, 1:]
+ preds_idx = preds.argmax(axis=2)
+ preds_prob = preds.max(axis=2)
+ text = self.decode(preds_idx, preds_prob, is_remove_duplicate=False)
+ if label is None:
+ return text
+ label = self.decode(label[:, 1:])
+ return text, label
+
+ def add_special_char(self, dict_character):
+ dict_character = ['', ''] + dict_character
+ return dict_character
+
+
+class ABINetLabelDecode(NRTRLabelDecode):
+ """ Convert between text-label and text-index """
+
+ def __init__(self, character_dict_path=None, use_space_char=False,
+ **kwargs):
+ super(ABINetLabelDecode, self).__init__(character_dict_path,
+ use_space_char)
+
+ def __call__(self, preds, label=None, *args, **kwargs):
+ if isinstance(preds, dict):
+ preds = preds['align'][-1].numpy()
+ elif isinstance(preds, paddle.Tensor):
+ preds = preds.numpy()
+ else:
+ preds = preds
+
+ preds_idx = preds.argmax(axis=2)
+ preds_prob = preds.max(axis=2)
+ text = self.decode(preds_idx, preds_prob, is_remove_duplicate=False)
+ if label is None:
+ return text
+ label = self.decode(label)
+ return text, label
+
+ def add_special_char(self, dict_character):
+ dict_character = [''] + dict_character
+ return dict_character
diff --git a/ppocr/utils/save_load.py b/ppocr/utils/save_load.py
index b09f1db6e9..3647111fdd 100644
--- a/ppocr/utils/save_load.py
+++ b/ppocr/utils/save_load.py
@@ -177,9 +177,9 @@ def save_model(model,
model.backbone.model.save_pretrained(model_prefix)
metric_prefix = os.path.join(model_prefix, 'metric')
# save metric and config
+ with open(metric_prefix + '.states', 'wb') as f:
+ pickle.dump(kwargs, f, protocol=2)
if is_best:
- with open(metric_prefix + '.states', 'wb') as f:
- pickle.dump(kwargs, f, protocol=2)
logger.info('save best model is to {}'.format(model_prefix))
else:
logger.info("save model in {}".format(model_prefix))
diff --git a/ppstructure/table/README.md b/ppstructure/table/README.md
index d21ef4aa38..b6804c6f09 100644
--- a/ppstructure/table/README.md
+++ b/ppstructure/table/README.md
@@ -18,7 +18,7 @@ The table recognition mainly contains three models
The table recognition flow chart is as follows
-
+
1. The coordinates of single-line text is detected by DB model, and then sends it to the recognition model to get the recognition result.
2. The table structure and cell coordinates is predicted by RARE model.
diff --git a/ppstructure/table/predict_table.py b/ppstructure/table/predict_table.py
index 402d6c2418..aa05459589 100644
--- a/ppstructure/table/predict_table.py
+++ b/ppstructure/table/predict_table.py
@@ -28,6 +28,7 @@ import numpy as np
import time
import tools.infer.predict_rec as predict_rec
import tools.infer.predict_det as predict_det
+import tools.infer.utility as utility
from ppocr.utils.utility import get_image_file_list, check_and_read_gif
from ppocr.utils.logging import get_logger
from ppstructure.table.matcher import distance, compute_iou
@@ -59,11 +60,37 @@ class TableSystem(object):
self.text_recognizer = predict_rec.TextRecognizer(
args) if text_recognizer is None else text_recognizer
self.table_structurer = predict_strture.TableStructurer(args)
+ self.benchmark = args.benchmark
+ self.predictor, self.input_tensor, self.output_tensors, self.config = utility.create_predictor(
+ args, 'table', logger)
+ if args.benchmark:
+ import auto_log
+ pid = os.getpid()
+ gpu_id = utility.get_infer_gpuid()
+ self.autolog = auto_log.AutoLogger(
+ model_name="table",
+ model_precision=args.precision,
+ batch_size=1,
+ data_shape="dynamic",
+ save_path=None, #args.save_log_path,
+ inference_config=self.config,
+ pids=pid,
+ process_name=None,
+ gpu_ids=gpu_id if args.use_gpu else None,
+ time_keys=[
+ 'preprocess_time', 'inference_time', 'postprocess_time'
+ ],
+ warmup=0,
+ logger=logger)
def __call__(self, img, return_ocr_result_in_table=False):
result = dict()
ori_im = img.copy()
+ if self.benchmark:
+ self.autolog.times.start()
structure_res, elapse = self.table_structurer(copy.deepcopy(img))
+ if self.benchmark:
+ self.autolog.times.stamp()
dt_boxes, elapse = self.text_detector(copy.deepcopy(img))
dt_boxes = sorted_boxes(dt_boxes)
if return_ocr_result_in_table:
@@ -77,13 +104,11 @@ class TableSystem(object):
box = [x_min, y_min, x_max, y_max]
r_boxes.append(box)
dt_boxes = np.array(r_boxes)
-
logger.debug("dt_boxes num : {}, elapse : {}".format(
len(dt_boxes), elapse))
if dt_boxes is None:
return None, None
img_crop_list = []
-
for i in range(len(dt_boxes)):
det_box = dt_boxes[i]
x0, y0, x1, y1 = expand(2, det_box, ori_im.shape)
@@ -92,10 +117,14 @@ class TableSystem(object):
rec_res, elapse = self.text_recognizer(img_crop_list)
logger.debug("rec_res num : {}, elapse : {}".format(
len(rec_res), elapse))
+ if self.benchmark:
+ self.autolog.times.stamp()
if return_ocr_result_in_table:
result['rec_res'] = rec_res
pred_html, pred = self.rebuild_table(structure_res, dt_boxes, rec_res)
result['html'] = pred_html
+ if self.benchmark:
+ self.autolog.times.end(stamp=True)
return result
def rebuild_table(self, structure_res, dt_boxes, rec_res):
@@ -213,6 +242,8 @@ def main(args):
logger.info('excel saved to {}'.format(excel_path))
elapse = time.time() - starttime
logger.info("Predict time : {:.3f}s".format(elapse))
+ if args.benchmark:
+ text_sys.autolog.report()
if __name__ == "__main__":
diff --git a/ppstructure/vqa/README_ch.md b/ppstructure/vqa/README_ch.md
index ff513f8f7d..b677dc07bc 100644
--- a/ppstructure/vqa/README_ch.md
+++ b/ppstructure/vqa/README_ch.md
@@ -52,7 +52,7 @@ PP-Structure 里的 DOC-VQA算法基于PaddleNLP自然语言处理算法库进
### 3.1 SER
- | 
+ | 
---|---
图中不同颜色的框表示不同的类别，对于XFUND数据集，有`QUESTION`, `ANSWER`, `HEADER` 3种类别
@@ -65,7 +65,7 @@ PP-Structure 里的 DOC-VQA算法基于PaddleNLP自然语言处理算法库进
### 3.2 RE
- | 
+ | 
---|---
diff --git a/test_tipc/build_server.sh b/test_tipc/build_server.sh
new file mode 100644
index 0000000000..3173359785
--- /dev/null
+++ b/test_tipc/build_server.sh
@@ -0,0 +1,69 @@
+#使用镜像：
+#registry.baidubce.com/paddlepaddle/paddle:latest-dev-cuda10.1-cudnn7-gcc82
+
+#编译Serving Server：
+
+#client和app可以直接使用release版本
+
+#server因为加入了自定义OP，需要重新编译
+
+apt-get update
+apt install -y libcurl4-openssl-dev libbz2-dev
+wget https://paddle-serving.bj.bcebos.com/others/centos_ssl.tar && tar xf centos_ssl.tar && rm -rf centos_ssl.tar && mv libcrypto.so.1.0.2k /usr/lib/libcrypto.so.1.0.2k && mv libssl.so.1.0.2k /usr/lib/libssl.so.1.0.2k && ln -sf /usr/lib/libcrypto.so.1.0.2k /usr/lib/libcrypto.so.10 && ln -sf /usr/lib/libssl.so.1.0.2k /usr/lib/libssl.so.10 && ln -sf /usr/lib/libcrypto.so.10 /usr/lib/libcrypto.so && ln -sf /usr/lib/libssl.so.10 /usr/lib/libssl.so
+
+# 安装go依赖
+rm -rf /usr/local/go
+wget -qO- https://paddle-ci.cdn.bcebos.com/go1.17.2.linux-amd64.tar.gz | tar -xz -C /usr/local
+export GOROOT=/usr/local/go
+export GOPATH=/root/gopath
+export PATH=$PATH:$GOPATH/bin:$GOROOT/bin
+go env -w GO111MODULE=on
+go env -w GOPROXY=https://goproxy.cn,direct
+go install github.com/grpc-ecosystem/grpc-gateway/protoc-gen-grpc-gateway@v1.15.2
+go install github.com/grpc-ecosystem/grpc-gateway/protoc-gen-swagger@v1.15.2
+go install github.com/golang/protobuf/protoc-gen-go@v1.4.3
+go install google.golang.org/grpc@v1.33.0
+go env -w GO111MODULE=auto
+
+# 下载opencv库
+wget https://paddle-qa.bj.bcebos.com/PaddleServing/opencv3.tar.gz && tar -xvf opencv3.tar.gz && rm -rf opencv3.tar.gz
+export OPENCV_DIR=$PWD/opencv3
+
+# clone Serving
+git clone https://github.com/PaddlePaddle/Serving.git -b develop --depth=1
+cd Serving
+export Serving_repo_path=$PWD
+git submodule update --init --recursive
+python -m pip install -r python/requirements.txt
+
+
+export PYTHON_INCLUDE_DIR=$(python -c "from distutils.sysconfig import get_python_inc; print(get_python_inc())")
+export PYTHON_LIBRARIES=$(python -c "import distutils.sysconfig as sysconfig; print(sysconfig.get_config_var('LIBDIR'))")
+export PYTHON_EXECUTABLE=`which python`
+
+export CUDA_PATH='/usr/local/cuda'
+export CUDNN_LIBRARY='/usr/local/cuda/lib64/'
+export CUDA_CUDART_LIBRARY='/usr/local/cuda/lib64/'
+export TENSORRT_LIBRARY_PATH='/usr/local/TensorRT6-cuda10.1-cudnn7/targets/x86_64-linux-gnu/'
+
+# cp 自定义OP代码
+cp -rf ../deploy/pdserving/general_detection_op.cpp ${Serving_repo_path}/core/general-server/op
+
+# 编译Server, export SERVING_BIN
+mkdir server-build-gpu-opencv && cd server-build-gpu-opencv
+cmake -DPYTHON_INCLUDE_DIR=$PYTHON_INCLUDE_DIR \
+ -DPYTHON_LIBRARIES=$PYTHON_LIBRARIES \
+ -DPYTHON_EXECUTABLE=$PYTHON_EXECUTABLE \
+ -DCUDA_TOOLKIT_ROOT_DIR=${CUDA_PATH} \
+ -DCUDNN_LIBRARY=${CUDNN_LIBRARY} \
+ -DCUDA_CUDART_LIBRARY=${CUDA_CUDART_LIBRARY} \
+ -DTENSORRT_ROOT=${TENSORRT_LIBRARY_PATH} \
+ -DOPENCV_DIR=${OPENCV_DIR} \
+ -DWITH_OPENCV=ON \
+ -DSERVER=ON \
+ -DWITH_GPU=ON ..
+make -j32
+
+python -m pip install python/dist/paddle*
+export SERVING_BIN=$PWD/core/general-server/serving
+cd ../../
diff --git a/test_tipc/common_func.sh b/test_tipc/common_func.sh
index 85dfe21725..f7d8a1e04a 100644
--- a/test_tipc/common_func.sh
+++ b/test_tipc/common_func.sh
@@ -57,10 +57,11 @@ function status_check(){
last_status=$1 # the exit code
run_command=$2
run_log=$3
+ model_name=$4
if [ $last_status -eq 0 ]; then
- echo -e "\033[33m Run successfully with command - ${run_command}! \033[0m" | tee -a ${run_log}
+ echo -e "\033[33m Run successfully with command - ${model_name} - ${run_command}! \033[0m" | tee -a ${run_log}
else
- echo -e "\033[33m Run failed with command - ${run_command}! \033[0m" | tee -a ${run_log}
+ echo -e "\033[33m Run failed with command - ${model_name} - ${run_command}! \033[0m" | tee -a ${run_log}
fi
}
diff --git a/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..a0c49a0812
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_PP-OCRv2
+use_opencv:True
+infer_model:./inference/ch_PP-OCRv2_det_infer/
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt --rec_img_h=32
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+--rec_model_dir:./inference/ch_PP-OCRv2_rec_infer/
+--benchmark:True
+--det:True
+--rec:True
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
index fcac6e3984..32b290a9ed 100644
--- a/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
@@ -6,10 +6,10 @@ infer_export:null
infer_quant:False
inference:tools/infer/predict_system.py
--use_gpu:False|True
---enable_mkldnn:False|True
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
+--use_tensorrt:False
--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
diff --git a/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt b/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
new file mode 100644
index 0000000000..24eb620eeb
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
@@ -0,0 +1,17 @@
+===========================paddle2onnx_params===========================
+model_name:ch_PP-OCRv2
+python:python3.7
+2onnx: paddle2onnx
+--det_model_dir:./inference/ch_PP-OCRv2_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_save_file:./inference/det_v2_onnx/model.onnx
+--rec_model_dir:./inference/ch_PP-OCRv2_rec_infer/
+--rec_save_file:./inference/rec_v2_onnx/model.onnx
+--opset_version:10
+--enable_onnx_checker:True
+inference:tools/infer/predict_system.py --rec_image_shape="3,32,320"
+--use_gpu:True|False
+--det_model_dir:
+--rec_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/00008790.jpg
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..f0456b5c35
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv2
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv2_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v2_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v2_client/
+--rec_dirname:./inference/ch_PP-OCRv2_rec_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v2_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v2_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..4ad64db03c
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv2
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv2_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v2_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v2_client/
+--rec_dirname:./inference/ch_PP-OCRv2_rec_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v2_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v2_client/
+serving_dir:./deploy/pdserving
+web_service:web_service.py --config=config.yml --opt op.det.concurrency="1" op.rec.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv2_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..7eccbd725c
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_PP-OCRv2_det
+use_opencv:True
+infer_model:./inference/ch_PP-OCRv2_det_infer/
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+--det:True
+--rec:False
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv2_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
new file mode 100644
index 0000000000..2e7906076f
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
@@ -0,0 +1,17 @@
+===========================paddle2onnx_params===========================
+model_name:ch_PP-OCRv2_det
+python:python3.7
+2onnx: paddle2onnx
+--det_model_dir:./inference/ch_PP-OCRv2_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_save_file:./inference/det_v2_onnx/model.onnx
+--rec_model_dir:
+--rec_save_file:
+--opset_version:10
+--enable_onnx_checker:True
+inference:tools/infer/predict_det.py
+--use_gpu:True|False
+--det_model_dir:
+--rec_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv2_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..587a7d7ea6
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv2_det
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv2_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v2_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v2_client/
+--rec_dirname:null
+--rec_serving_server:null
+--rec_serving_client:null
+serving_dir:./deploy/pdserving
+web_service:web_service_det.py --config=config.yml --opt op.det.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv2_det/train_infer_python.txt b/test_tipc/configs/ch_PP-OCRv2_det/train_infer_python.txt
index 797cf53a1d..cab0cb0aa3 100644
--- a/test_tipc/configs/ch_PP-OCRv2_det/train_infer_python.txt
+++ b/test_tipc/configs/ch_PP-OCRv2_det/train_infer_python.txt
@@ -1,10 +1,10 @@
===========================train_params===========================
-model_name:ch_PPOCRv2_det
+model_name:ch_PP-OCRv2_det
python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:fp32
-Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=500
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:null
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/ch_PP-OCRv2_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..91a6288eb0
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:ch_PP-OCRv2_det
+python:python3.7
+gpu_list:192.168.0.1,192.168.0.2;0,1
+Global.use_gpu:True
+Global.auto_cast:fp32
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./train_data/icdar2015/text_localization/ch4_test_images/
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c configs/det/ch_PP-OCRv2/ch_PP-OCRv2_det_cml.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:null
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c configs/det/ch_PP-OCRv2/ch_PP-OCRv2_det_cml.yml -o
+quant_export:null
+fpgm_export:
+distill_export:null
+export1:null
+export2:null
+inference_dir:Student
+infer_model:./inference/ch_PP-OCRv2_det_infer/
+infer_export:null
+infer_quant:False
+inference:tools/infer/predict_det.py
+--use_gpu:False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,640,640]}];[{float32,[3,960,960]}]
diff --git a/test_tipc/configs/ch_PP-OCRv2_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 033d40a80a..85b0ebcb9d 100644
--- a/test_tipc/configs/ch_PP-OCRv2_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_PP-OCRv2_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -1,10 +1,10 @@
===========================train_params===========================
-model_name:ch_PPOCRv2_det
+model_name:ch_PP-OCRv2_det
python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:amp
-Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=500
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:null
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/ch_PP-OCRv2_det_PACT/train_infer_python.txt b/test_tipc/configs/ch_PP-OCRv2_det/train_pact_infer_python.txt
similarity index 91%
rename from test_tipc/configs/ch_PP-OCRv2_det_PACT/train_infer_python.txt
rename to test_tipc/configs/ch_PP-OCRv2_det/train_pact_infer_python.txt
index 038fa85061..1a20f97fdb 100644
--- a/test_tipc/configs/ch_PP-OCRv2_det_PACT/train_infer_python.txt
+++ b/test_tipc/configs/ch_PP-OCRv2_det/train_pact_infer_python.txt
@@ -1,10 +1,10 @@
===========================train_params===========================
-model_name:ch_PPOCRv2_det_PACT
+model_name:ch_PP-OCRv2_det_PACT
python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:fp32
-Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=500
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=1|whole_train_whole_infer=4
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:null
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_det/train_ptq_infer_python.txt
similarity index 84%
rename from test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
rename to test_tipc/configs/ch_PP-OCRv2_det/train_ptq_infer_python.txt
index 1aad65b687..ccc9e5ced0 100644
--- a/test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_PP-OCRv2_det/train_ptq_infer_python.txt
@@ -1,5 +1,5 @@
===========================kl_quant_params===========================
-model_name:PPOCRv2_ocr_det_kl
+model_name:ch_PP-OCRv2_det_KL
python:python3.7
Global.pretrained_model:null
Global.save_inference_dir:null
@@ -8,10 +8,10 @@ infer_export:deploy/slim/quantization/quant_kl.py -c configs/det/ch_PP-OCRv2/ch_
infer_quant:True
inference:tools/infer/predict_det.py
--use_gpu:False|True
---enable_mkldnn:True
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
+--use_tensorrt:False
--precision:int8
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
diff --git a/test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..1975e099d7
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_PP-OCRv2_det_KL
+use_opencv:True
+infer_model:./inference/ch_PP-OCRv2_det_klquant_infer
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+--det:True
+--rec:False
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..e306b0a92a
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv2_det_KL
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv2_det_klquant_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v2_kl_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v2_kl_client/
+--rec_dirname:./inference/ch_PP-OCRv2_rec_klquant_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v2_kl_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v2_kl_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..2c96d2bfd8
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv2_det_KL
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv2_det_klquant_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v2_kl_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v2_kl_client/
+--rec_dirname:null
+--rec_serving_server:null
+--rec_serving_client:null
+serving_dir:./deploy/pdserving
+web_service:web_service_det.py --config=config.yml --opt op.det.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv2_det_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_det_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..43ef97d506
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_det_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_PP-OCRv2_det_PACT
+use_opencv:True
+infer_model:./inference/ch_PP-OCRv2_det_pact_infer
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+--det:True
+--rec:False
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv2_det_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_det_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..b2d929b99c
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_det_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv2_det_PACT
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv2_det_pact_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v2_pact_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v2_pact_client/
+--rec_dirname:./inference/ch_PP-OCRv2_rec_pact_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v2_pact_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v2_pact_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv2_det_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_det_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..d5d99ab56d
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_det_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv2_det_PACT
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv2_det_pact_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v2_pact_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v2_pact_client/
+--rec_dirname:null
+--rec_serving_server:null
+--rec_serving_client:null
+serving_dir:./deploy/pdserving
+web_service:web_service_det.py --config=config.yml --opt op.det.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..b1bff00b09
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_PP-OCRv2_rec
+use_opencv:True
+infer_model:./inference/ch_PP-OCRv2_rec_infer/
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt --rec_img_h=32
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference/
+null:null
+--benchmark:True
+--det:False
+--rec:True
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
new file mode 100644
index 0000000000..e374a5d821
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
@@ -0,0 +1,17 @@
+===========================paddle2onnx_params===========================
+model_name:ch_PP-OCRv2_rec
+python:python3.7
+2onnx: paddle2onnx
+--det_model_dir:
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_save_file:
+--rec_model_dir:./inference/ch_PP-OCRv2_rec_infer/
+--rec_save_file:./inference/rec_v2_onnx/model.onnx
+--opset_version:10
+--enable_onnx_checker:True
+inference:tools/infer/predict_rec.py --rec_image_shape="3,32,320"
+--use_gpu:True|False
+--det_model_dir:
+--rec_model_dir:
+--image_dir:./inference/rec_inference/
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..e9e90d372e
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv2_rec
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:null
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:null
+--det_serving_client:null
+--rec_dirname:./inference/ch_PP-OCRv2_rec_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v2_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v2_client/
+serving_dir:./deploy/pdserving
+web_service:web_service_rec.py --config=config.yml --opt op.rec.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py --det=False
+--image_dir:../../inference/rec_inference
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec/train_infer_python.txt b/test_tipc/configs/ch_PP-OCRv2_rec/train_infer_python.txt
index 188eb3ccc5..df42b342ba 100644
--- a/test_tipc/configs/ch_PP-OCRv2_rec/train_infer_python.txt
+++ b/test_tipc/configs/ch_PP-OCRv2_rec/train_infer_python.txt
@@ -1,10 +1,10 @@
===========================train_params===========================
-model_name:PPOCRv2_ocr_rec
+model_name:ch_PP-OCRv2_rec
python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:fp32
-Global.epoch_num:lite_train_lite_infer=3|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=3|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=128
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:null
infer_quant:False
inference:tools/infer/predict_rec.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:False|True
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
null:null
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..5795bc27e6
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:ch_PP-OCRv2_rec
+python:python3.7
+gpu_list:192.168.0.1,192.168.0.2;0,1
+Global.use_gpu:True
+Global.auto_cast:fp32
+Global.epoch_num:lite_train_lite_infer=3|whole_train_whole_infer=50
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=128
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./inference/rec_inference
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c test_tipc/configs/ch_PP-OCRv2_rec/ch_PP-OCRv2_rec_distillation.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:null
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c test_tipc/configs/ch_PP-OCRv2_rec/ch_PP-OCRv2_rec_distillation.yml -o
+quant_export:
+fpgm_export:
+distill_export:null
+export1:null
+export2:null
+inference_dir:Student
+infer_model:./inference/ch_PP-OCRv2_rec_infer
+infer_export:null
+infer_quant:False
+inference:tools/infer/predict_rec.py
+--use_gpu:False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1|6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference
+null:null
+--benchmark:True
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,32,320]}]
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 7c438cb8a3..1b8800f5cf 100644
--- a/test_tipc/configs/ch_PP-OCRv2_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_PP-OCRv2_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -1,10 +1,10 @@
===========================train_params===========================
-model_name:PPOCRv2_ocr_rec
+model_name:ch_PP-OCRv2_rec
python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:amp
-Global.epoch_num:lite_train_lite_infer=3|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=3|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=128
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:null
infer_quant:False
inference:tools/infer/predict_rec.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:False|True
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
null:null
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec_PACT/train_infer_python.txt b/test_tipc/configs/ch_PP-OCRv2_rec/train_pact_infer_python.txt
similarity index 88%
rename from test_tipc/configs/ch_PP-OCRv2_rec_PACT/train_infer_python.txt
rename to test_tipc/configs/ch_PP-OCRv2_rec/train_pact_infer_python.txt
index 98c125229d..0ac75eff07 100644
--- a/test_tipc/configs/ch_PP-OCRv2_rec_PACT/train_infer_python.txt
+++ b/test_tipc/configs/ch_PP-OCRv2_rec/train_pact_infer_python.txt
@@ -1,10 +1,10 @@
===========================train_params===========================
-model_name:ch_PPOCRv2_rec_PACT
+model_name:ch_PP-OCRv2_rec_PACT
python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:fp32
-Global.epoch_num:lite_train_lite_infer=6|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=128
Global.pretrained_model:pretrain_models/ch_PP-OCRv2_rec_train/best_accuracy
@@ -39,11 +39,11 @@ infer_export:null
infer_quant:True
inference:tools/infer/predict_rec.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:False|True
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
null:null
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_rec/train_ptq_infer_python.txt
similarity index 76%
rename from test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
rename to test_tipc/configs/ch_PP-OCRv2_rec/train_ptq_infer_python.txt
index 083a3ae26e..c30e0858ef 100644
--- a/test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_PP-OCRv2_rec/train_ptq_infer_python.txt
@@ -1,17 +1,17 @@
===========================kl_quant_params===========================
-model_name:PPOCRv2_ocr_rec_kl
+model_name:ch_PP-OCRv2_rec_KL
python:python3.7
Global.pretrained_model:null
Global.save_inference_dir:null
infer_model:./inference/ch_PP-OCRv2_rec_infer/
infer_export:deploy/slim/quantization/quant_kl.py -c test_tipc/configs/ch_PP-OCRv2_rec/ch_PP-OCRv2_rec_distillation.yml -o
infer_quant:True
-inference:tools/infer/predict_rec.py
+inference:tools/infer/predict_rec.py --rec_image_shape="3,32,320"
--use_gpu:False|True
---enable_mkldnn:False|True
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True
+--use_tensorrt:False
--precision:int8
--rec_model_dir:
--image_dir:./inference/rec_inference
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..95e4062d14
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_PP-OCRv2_rec_KL
+use_opencv:True
+infer_model:./inference/ch_PP-OCRv2_rec_klquant_infer
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt --rec_img_h=32
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference/
+null:null
+--benchmark:True
+--det:False
+--rec:True
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..34d4007b15
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv2_rec_KL
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv2_det_klquant_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v2_kl_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v2_kl_client/
+--rec_dirname:./inference/ch_PP-OCRv2_rec_klquant_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v2_kl_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v2_kl_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..3405f2b587
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv2_rec_KL
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:null
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:null
+--det_serving_client:null
+--rec_dirname:./inference/ch_PP-OCRv2_rec_klquant_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v2_kl_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v2_kl_client/
+serving_dir:./deploy/pdserving
+web_service:web_service_rec.py --config=config.yml --opt op.rec.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py --det=False
+--image_dir:../../inference/rec_inference
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_rec_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..b807eadd3f
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_rec_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_PP-OCRv2_rec_PACT
+use_opencv:True
+infer_model:./inference/ch_PP-OCRv2_rec_pact_infer
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt --rec_img_h=32
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference/
+null:null
+--benchmark:True
+--det:False
+--rec:True
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_rec_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..2a174b9e31
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_rec_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv2_rec_PACT
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv2_det_pact_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v2_pact_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v2_pact_client/
+--rec_dirname:./inference/ch_PP-OCRv2_rec_pact_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v2_pact_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v2_pact_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv2_rec_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..2b7ed81172
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv2_rec_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv2_rec_PACT
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:null
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:null
+--det_serving_client:null
+--rec_dirname:./inference/ch_PP-OCRv2_rec_pact_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v2_pact_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v2_pact_client/
+serving_dir:./deploy/pdserving
+web_service:web_service_rec.py --config=config.yml --opt op.rec.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py --det=False
+--image_dir:../../inference/rec_inference
diff --git a/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..794af27d90
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_PP-OCRv3
+use_opencv:True
+infer_model:./inference/ch_PP-OCRv3_det_infer/
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr --rec_img_h=48 --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+--rec_model_dir:./inference/ch_PP-OCRv3_rec_infer/
+--benchmark:True
+--det:True
+--rec:True
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..afacdc1405
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================ch_PP-OCRv2===========================
+model_name:ch_PP-OCRv3
+python:python3.7
+infer_model:./inference/ch_PP-OCRv3_det_infer/
+infer_export:null
+infer_quant:False
+inference:tools/infer/predict_system.py --rec_image_shape="3,48,320"
+--use_gpu:False|True
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+--rec_model_dir:./inference/ch_PP-OCRv3_rec_infer/
+--benchmark:True
+null:null
+null:null
diff --git a/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_lite_cpp_arm_cpu.txt b/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_lite_cpp_arm_cpu.txt
new file mode 100644
index 0000000000..133a78c84f
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_lite_cpp_arm_cpu.txt
@@ -0,0 +1,13 @@
+===========================lite_params===========================
+inference:./ocr_db_crnn system
+runtime_device:ARM_CPU
+det_infer_model:ch_PP-OCRv3_det_infer|ch_PP-OCRv3_det_slim_quant_infer
+rec_infer_model:ch_PP-OCRv3_rec_infer|ch_PP-OCRv3_rec_slim_quant_infer
+cls_infer_model:ch_ppocr_mobile_v2.0_cls_infer|ch_ppocr_mobile_v2.0_cls_slim_infer
+--cpu_threads:1|4
+--det_batch_size:1
+--rec_batch_size:1
+--image_dir:./test_data/icdar2015_lite/text_localization/ch4_test_images/
+--config_dir:./config.txt
+--rec_dict_dir:./ppocr_keys_v1.txt
+--benchmark:True
diff --git a/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_lite_cpp_arm_gpu_opencl.txt b/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_lite_cpp_arm_gpu_opencl.txt
new file mode 100644
index 0000000000..86e49fc100
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_lite_cpp_arm_gpu_opencl.txt
@@ -0,0 +1,13 @@
+===========================lite_params===========================
+inference:./ocr_db_crnn system
+runtime_device:ARM_GPU_OPENCL
+det_infer_model:ch_PP-OCRv3_det_infer|ch_PP-OCRv3_det_slim_quant_infer
+rec_infer_model:ch_PP-OCRv3_rec_infer|ch_PP-OCRv3_rec_slim_quant_infer
+cls_infer_model:ch_ppocr_mobile_v2.0_cls_infer|ch_ppocr_mobile_v2.0_cls_slim_infer
+--cpu_threads:1|4
+--det_batch_size:1
+--rec_batch_size:1
+--image_dir:./test_data/icdar2015_lite/text_localization/ch4_test_images/
+--config_dir:./config.txt
+--rec_dict_dir:./ppocr_keys_v1.txt
+--benchmark:True
diff --git a/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt b/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
new file mode 100644
index 0000000000..bf2556ef17
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
@@ -0,0 +1,17 @@
+===========================paddle2onnx_params===========================
+model_name:ch_PP-OCRv3
+python:python3.7
+2onnx: paddle2onnx
+--det_model_dir:./inference/ch_PP-OCRv3_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_save_file:./inference/det_v3_onnx/model.onnx
+--rec_model_dir:./inference/ch_PP-OCRv3_rec_infer/
+--rec_save_file:./inference/rec_v3_onnx/model.onnx
+--opset_version:10
+--enable_onnx_checker:True
+inference:tools/infer/predict_system.py --rec_image_shape="3,48,320"
+--use_gpu:True|False
+--det_model_dir:
+--rec_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/00008790.jpg
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..91c57bed1b
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv3
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv3_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v3_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v3_client/
+--rec_dirname:./inference/ch_PP-OCRv3_rec_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v3_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v3_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..6f699ef5c0
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv3
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv3_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v3_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v3_client/
+--rec_dirname:./inference/ch_PP-OCRv3_rec_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v3_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v3_client/
+serving_dir:./deploy/pdserving
+web_service:web_service.py --config=config.yml --opt op.det.concurrency="1" op.rec.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..aecd0dd437
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_PP-OCRv3_det
+use_opencv:True
+infer_model:./inference/ch_PP-OCRv3_det_infer/
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+--det:True
+--rec:False
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_lite_cpp_arm_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_lite_cpp_arm_cpu.txt
new file mode 100644
index 0000000000..cbc101f93b
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_lite_cpp_arm_cpu.txt
@@ -0,0 +1,13 @@
+===========================lite_params===========================
+inference:./ocr_db_crnn det
+runtime_device:ARM_CPU
+det_infer_model:ch_PP-OCRv3_det_infer|ch_PP-OCRv3_det_slim_quant_infer
+null:null
+null:null
+--cpu_threads:1|4
+--det_batch_size:1
+null:null
+--image_dir:./test_data/icdar2015_lite/text_localization/ch4_test_images/
+--config_dir:./config.txt
+null:null
+--benchmark:True
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_lite_cpp_arm_gpu_opencl.txt b/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_lite_cpp_arm_gpu_opencl.txt
new file mode 100644
index 0000000000..ba3f5b71e5
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_lite_cpp_arm_gpu_opencl.txt
@@ -0,0 +1,13 @@
+===========================lite_params===========================
+inference:./ocr_db_crnn det
+runtime_device:ARM_GPU_OPENCL
+det_infer_model:ch_PP-OCRv3_det_infer|ch_PP-OCRv3_det_slim_quant_infer
+null:null
+null:null
+--cpu_threads:1|4
+--det_batch_size:1
+null:null
+--image_dir:./test_data/icdar2015_lite/text_localization/ch4_test_images/
+--config_dir:./config.txt
+null:null
+--benchmark:True
diff --git a/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
new file mode 100644
index 0000000000..a448713b1c
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
@@ -0,0 +1,17 @@
+===========================paddle2onnx_params===========================
+model_name:ch_PP-OCRv3_det
+python:python3.7
+2onnx: paddle2onnx
+--det_model_dir:./inference/ch_PP-OCRv3_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_save_file:./inference/det_v3_onnx/model.onnx
+--rec_model_dir:
+--rec_save_file:
+--opset_version:10
+--enable_onnx_checker:True
+inference:tools/infer/predict_det.py
+--use_gpu:True|False
+--det_model_dir:
+--rec_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..6e2ec22cbd
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv3_det
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv3_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v3_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v3_client/
+--rec_dirname:null
+--rec_serving_server:null
+--rec_serving_client:null
+serving_dir:./deploy/pdserving
+web_service:web_service_det.py --config=config.yml --opt op.det.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv3_det/train_infer_python.txt b/test_tipc/configs/ch_PP-OCRv3_det/train_infer_python.txt
new file mode 100644
index 0000000000..a69e0ab81e
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_det/train_infer_python.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:ch_PP-OCRv3_det
+python:python3.7
+gpu_list:0|0,1
+Global.use_gpu:True|True
+Global.auto_cast:fp32
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./train_data/icdar2015/text_localization/ch4_test_images/
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:null
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml -o
+quant_export:null
+fpgm_export:
+distill_export:null
+export1:null
+export2:null
+inference_dir:Student
+infer_model:./inference/ch_PP-OCRv3_det_infer/
+infer_export:null
+infer_quant:False
+inference:tools/infer/predict_det.py
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,640,640]}];[{float32,[3,960,960]}]
diff --git a/test_tipc/configs/ch_PP-OCRv3_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..7e987125a6
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:ch_PP-OCRv3_det
+python:python3.7
+gpu_list:192.168.0.1,192.168.0.2;0,1
+Global.use_gpu:True
+Global.auto_cast:fp32
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./train_data/icdar2015/text_localization/ch4_test_images/
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:null
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml -o
+quant_export:null
+fpgm_export:
+distill_export:null
+export1:null
+export2:null
+inference_dir:Student
+infer_model:./inference/ch_PP-OCRv3_det_infer/
+infer_export:null
+infer_quant:False
+inference:tools/infer/predict_det.py
+--use_gpu:False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,640,640]}];[{float32,[3,960,960]}]
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
similarity index 65%
rename from test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
rename to test_tipc/configs/ch_PP-OCRv3_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 1f9bec12ad..fe72cfb4e9 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_PP-OCRv3_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -1,10 +1,10 @@
===========================train_params===========================
-model_name:ch_ppocr_mobile_v2.0_det_PACT
+model_name:ch_PP-OCRv3_det
python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:amp
-Global.epoch_num:lite_train_lite_infer=20|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
Global.pretrained_model:null
@@ -12,9 +12,9 @@ train_model_name:latest
train_infer_img_dir:./train_data/icdar2015/text_localization/ch4_test_images/
null:null
##
-trainer:pact_train
-norm_train:null
-pact_train:deploy/slim/quantization/quant.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o
+trainer:norm_train
+norm_train:tools/train.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml -o
+pact_train:null
fpgm_train:null
distill_train:null
null:null
@@ -27,23 +27,23 @@ null:null
===========================infer_params===========================
Global.save_inference_dir:./output/
Global.checkpoints:
-norm_export:null
-quant_export:deploy/slim/quantization/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o
-fpgm_export:null
+norm_export:tools/export_model.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml -o
+quant_export:null
+fpgm_export:
distill_export:null
export1:null
export2:null
-inference_dir:null
-train_model:./inference/ch_ppocr_mobile_v2.0_det_prune_infer/
+inference_dir:Student
+infer_model:./inference/ch_PP-OCRv3_det_infer/
infer_export:null
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/ch_PP-OCRv2_det_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_det/train_pact_infer_python.txt
similarity index 78%
rename from test_tipc/configs/ch_PP-OCRv2_det_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
rename to test_tipc/configs/ch_PP-OCRv3_det/train_pact_infer_python.txt
index d922a4a5da..b536e69b05 100644
--- a/test_tipc/configs/ch_PP-OCRv2_det_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_PP-OCRv3_det/train_pact_infer_python.txt
@@ -1,12 +1,12 @@
===========================train_params===========================
-model_name:ch_PPOCRv2_det_PACT
+model_name:ch_PP-OCRv3_det_PACT
python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
-Global.auto_cast:amp
-Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=500
+Global.auto_cast:fp32
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
Global.save_model_dir:./output/
-Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
+Train.loader.batch_size_per_card:lite_train_lite_infer=1|whole_train_whole_infer=4
Global.pretrained_model:null
train_model_name:latest
train_infer_img_dir:./train_data/icdar2015/text_localization/ch4_test_images/
@@ -14,7 +14,7 @@ null:null
##
trainer:pact_train
norm_train:null
-pact_train:deploy/slim/quantization/quant.py -c configs/det/ch_PP-OCRv2/ch_PP-OCRv2_det_cml.yml -o
+pact_train:deploy/slim/quantization/quant.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml -o
fpgm_train:null
distill_train:null
null:null
@@ -28,22 +28,22 @@ null:null
Global.save_inference_dir:./output/
Global.checkpoints:
norm_export:null
-quant_export:deploy/slim/quantization/export_model.py -c configs/det/ch_PP-OCRv2/ch_PP-OCRv2_det_cml.yml -o
+quant_export:deploy/slim/quantization/export_model.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml -o
fpgm_export:
distill_export:null
export1:null
export2:null
inference_dir:Student
-infer_model:./inference/ch_PP-OCRv2_det_infer/
+infer_model:./inference/ch_PP-OCRv3_det_infer/
infer_export:null
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/ch_PP-OCRv3_det/train_ptq_infer_python.txt b/test_tipc/configs/ch_PP-OCRv3_det/train_ptq_infer_python.txt
new file mode 100644
index 0000000000..c27e08a640
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_det/train_ptq_infer_python.txt
@@ -0,0 +1,21 @@
+===========================kl_quant_params===========================
+model_name:ch_PP-OCRv3_det_KL
+python:python3.7
+Global.pretrained_model:null
+Global.save_inference_dir:null
+infer_model:./inference/ch_PP-OCRv3_det_infer/
+infer_export:deploy/slim/quantization/quant_kl.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml -o
+infer_quant:True
+inference:tools/infer/predict_det.py
+--use_gpu:False|True
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:int8
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+null:null
+null:null
diff --git a/test_tipc/configs/ch_PP-OCRv3_det_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_det_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..a34ffe22ae
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_det_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_PP-OCRv3_det_KL
+use_opencv:True
+infer_model:./inference/ch_PP-OCRv3_det_klquant_infer
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+--det:True
+--rec:False
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv3_det_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_det_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..22b429760c
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_det_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv3_det_KL
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv3_det_klquant_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v3_kl_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v3_kl_client/
+--rec_dirname:./inference/ch_PP-OCRv3_rec_klquant_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v3_kl_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v3_kl_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv3_det_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_det_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..23dbc49a38
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_det_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv3_det_KL
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv3_det_klquant_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v3_kl_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v3_kl_client/
+--rec_dirname:null
+--rec_serving_server:null
+--rec_serving_client:null
+serving_dir:./deploy/pdserving
+web_service:web_service_det.py --config=config.yml --opt op.det.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv3_det_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_det_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..3198b87552
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_det_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_PP-OCRv3_det_PACT
+use_opencv:True
+infer_model:./inference/ch_PP-OCRv3_det_pact_infer
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+--det:True
+--rec:False
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv3_det_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_det_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..7d300f4456
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_det_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv3_det_PACT
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv3_det_pact_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v3_pact_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v3_pact_client/
+--rec_dirname:./inference/ch_PP-OCRv3_rec_pact_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v3_pact_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v3_pact_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv3_det_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_det_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..4546644cbc
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_det_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv3_det_PACT
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv3_det_pact_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v3_pact_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v3_pact_client/
+--rec_dirname:null
+--rec_serving_server:null
+--rec_serving_client:null
+serving_dir:./deploy/pdserving
+web_service:web_service_det.py --config=config.yml --opt op.det.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml b/test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml
new file mode 100644
index 0000000000..f704a1dfb5
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml
@@ -0,0 +1,205 @@
+Global:
+ debug: false
+ use_gpu: true
+ epoch_num: 800
+ log_smooth_window: 20
+ print_batch_step: 10
+ save_model_dir: ./output/rec_ppocr_v3_distillation
+ save_epoch_step: 3
+ eval_batch_step: [0, 2000]
+ cal_metric_during_train: true
+ pretrained_model:
+ checkpoints:
+ save_inference_dir:
+ use_visualdl: false
+ infer_img: doc/imgs_words/ch/word_1.jpg
+ character_dict_path: ppocr/utils/ppocr_keys_v1.txt
+ max_text_length: &max_text_length 25
+ infer_mode: false
+ use_space_char: true
+ distributed: true
+ save_res_path: ./output/rec/predicts_ppocrv3_distillation.txt
+
+
+Optimizer:
+ name: Adam
+ beta1: 0.9
+ beta2: 0.999
+ lr:
+ name: Piecewise
+ decay_epochs : [700, 800]
+ values : [0.0005, 0.00005]
+ warmup_epoch: 5
+ regularizer:
+ name: L2
+ factor: 3.0e-05
+
+
+Architecture:
+ model_type: &model_type "rec"
+ name: DistillationModel
+ algorithm: Distillation
+ Models:
+ Teacher:
+ pretrained:
+ freeze_params: false
+ return_all_feats: true
+ model_type: *model_type
+ algorithm: SVTR
+ Transform:
+ Backbone:
+ name: MobileNetV1Enhance
+ scale: 0.5
+ last_conv_stride: [1, 2]
+ last_pool_type: avg
+ Head:
+ name: MultiHead
+ head_list:
+ - CTCHead:
+ Neck:
+ name: svtr
+ dims: 64
+ depth: 2
+ hidden_dims: 120
+ use_guide: True
+ Head:
+ fc_decay: 0.00001
+ - SARHead:
+ enc_dim: 512
+ max_text_length: *max_text_length
+ Student:
+ pretrained:
+ freeze_params: false
+ return_all_feats: true
+ model_type: *model_type
+ algorithm: SVTR
+ Transform:
+ Backbone:
+ name: MobileNetV1Enhance
+ scale: 0.5
+ last_conv_stride: [1, 2]
+ last_pool_type: avg
+ Head:
+ name: MultiHead
+ head_list:
+ - CTCHead:
+ Neck:
+ name: svtr
+ dims: 64
+ depth: 2
+ hidden_dims: 120
+ use_guide: True
+ Head:
+ fc_decay: 0.00001
+ - SARHead:
+ enc_dim: 512
+ max_text_length: *max_text_length
+Loss:
+ name: CombinedLoss
+ loss_config_list:
+ - DistillationDMLLoss:
+ weight: 1.0
+ act: "softmax"
+ use_log: true
+ model_name_pairs:
+ - ["Student", "Teacher"]
+ key: head_out
+ multi_head: True
+ dis_head: ctc
+ name: dml_ctc
+ - DistillationDMLLoss:
+ weight: 0.5
+ act: "softmax"
+ use_log: true
+ model_name_pairs:
+ - ["Student", "Teacher"]
+ key: head_out
+ multi_head: True
+ dis_head: sar
+ name: dml_sar
+ - DistillationDistanceLoss:
+ weight: 1.0
+ mode: "l2"
+ model_name_pairs:
+ - ["Student", "Teacher"]
+ key: backbone_out
+ - DistillationCTCLoss:
+ weight: 1.0
+ model_name_list: ["Student", "Teacher"]
+ key: head_out
+ multi_head: True
+ - DistillationSARLoss:
+ weight: 1.0
+ model_name_list: ["Student", "Teacher"]
+ key: head_out
+ multi_head: True
+
+PostProcess:
+ name: DistillationCTCLabelDecode
+ model_name: ["Student", "Teacher"]
+ key: head_out
+ multi_head: True
+
+Metric:
+ name: DistillationMetric
+ base_metric_name: RecMetric
+ main_indicator: acc
+ key: "Student"
+ ignore_space: True
+
+Train:
+ dataset:
+ name: SimpleDataSet
+ data_dir: ./train_data/ic15_data/
+ ext_op_transform_idx: 1
+ label_file_list:
+ - ./train_data/ic15_data/rec_gt_train_lite.txt
+ transforms:
+ - DecodeImage:
+ img_mode: BGR
+ channel_first: false
+ - RecConAug:
+ prob: 0.5
+ ext_data_num: 2
+ image_shape: [48, 320, 3]
+ - RecAug:
+ - MultiLabelEncode:
+ - RecResizeImg:
+ image_shape: [3, 48, 320]
+ - KeepKeys:
+ keep_keys:
+ - image
+ - label_ctc
+ - label_sar
+ - length
+ - valid_ratio
+ loader:
+ shuffle: true
+ batch_size_per_card: 128
+ drop_last: true
+ num_workers: 4
+Eval:
+ dataset:
+ name: SimpleDataSet
+ data_dir: ./train_data/ic15_data
+ label_file_list:
+ - ./train_data/ic15_data/rec_gt_test_lite.txt
+ transforms:
+ - DecodeImage:
+ img_mode: BGR
+ channel_first: false
+ - MultiLabelEncode:
+ - RecResizeImg:
+ image_shape: [3, 48, 320]
+ - KeepKeys:
+ keep_keys:
+ - image
+ - label_ctc
+ - label_sar
+ - length
+ - valid_ratio
+ loader:
+ shuffle: false
+ drop_last: false
+ batch_size_per_card: 128
+ num_workers: 4
diff --git a/test_tipc/configs/ch_PP-OCRv3_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..9d6ca2cf5e
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_PP-OCRv3_rec
+use_opencv:True
+infer_model:./inference/ch_PP-OCRv3_rec_infer/
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr --rec_img_h=48 --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference/
+null:null
+--benchmark:True
+--det:False
+--rec:True
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv3_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
new file mode 100644
index 0000000000..9114c0acfd
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
@@ -0,0 +1,17 @@
+===========================paddle2onnx_params===========================
+model_name:ch_PP-OCRv3_rec
+python:python3.7
+2onnx: paddle2onnx
+--det_model_dir:
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_save_file:
+--rec_model_dir:./inference/ch_PP-OCRv3_rec_infer/
+--rec_save_file:./inference/rec_v3_onnx/model.onnx
+--opset_version:10
+--enable_onnx_checker:True
+inference:tools/infer/predict_rec.py --rec_image_shape="3,48,320"
+--use_gpu:True|False
+--det_model_dir:
+--rec_model_dir:
+--image_dir:./inference/rec_inference/
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv3_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..f01db2e950
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv3_rec
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:null
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:null
+--det_serving_client:null
+--rec_dirname:./inference/ch_PP-OCRv3_rec_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v3_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v3_client/
+serving_dir:./deploy/pdserving
+web_service:web_service_rec.py --config=config.yml --opt op.rec.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py --det=False
+--image_dir:../../inference/rec_inference
diff --git a/test_tipc/configs/ch_PP-OCRv3_rec/train_infer_python.txt b/test_tipc/configs/ch_PP-OCRv3_rec/train_infer_python.txt
new file mode 100644
index 0000000000..1feb9d49fc
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_rec/train_infer_python.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:ch_PP-OCRv3_rec
+python:python3.7
+gpu_list:0|0,1
+Global.use_gpu:True|True
+Global.auto_cast:fp32
+Global.epoch_num:lite_train_lite_infer=3|whole_train_whole_infer=50
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=128
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./inference/rec_inference
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:null
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml -o
+quant_export:
+fpgm_export:
+distill_export:null
+export1:null
+export2:null
+inference_dir:Student
+infer_model:./inference/ch_PP-OCRv3_rec_infer
+infer_export:null
+infer_quant:False
+inference:tools/infer/predict_rec.py --rec_image_shape="3,48,320"
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1|6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference
+null:null
+--benchmark:True
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,48,320]}]
diff --git a/test_tipc/configs/ch_PP-OCRv3_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..7fcc8b4418
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:ch_PP-OCRv3_rec
+python:python3.7
+gpu_list:192.168.0.1,192.168.0.2;0,1
+Global.use_gpu:True
+Global.auto_cast:fp32
+Global.epoch_num:lite_train_lite_infer=3|whole_train_whole_infer=50
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=64
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./inference/rec_inference
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:null
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml -o
+quant_export:
+fpgm_export:
+distill_export:null
+export1:null
+export2:null
+inference_dir:Student
+infer_model:./inference/ch_PP-OCRv3_rec_infer
+infer_export:null
+infer_quant:False
+inference:tools/infer/predict_rec.py --rec_image_shape="3,48,320"
+--use_gpu:False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1|6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference
+null:null
+--benchmark:True
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,48,320]}]
diff --git a/test_tipc/configs/ch_PP-OCRv3_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..99e3c42477
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:ch_PP-OCRv3_rec
+python:python3.7
+gpu_list:0|0,1
+Global.use_gpu:True|True
+Global.auto_cast:amp
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=128
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./inference/rec_inference
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:null
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml -o
+quant_export:
+fpgm_export:
+distill_export:null
+export1:null
+export2:null
+inference_dir:Student
+infer_model:./inference/ch_PP-OCRv3_rec_infer
+infer_export:null
+infer_quant:False
+inference:tools/infer/predict_rec.py --rec_image_shape="3,48,320"
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference
+null:null
+--benchmark:True
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,48,320]}]
diff --git a/test_tipc/configs/ch_PP-OCRv2_rec_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_rec/train_pact_infer_python.txt
similarity index 66%
rename from test_tipc/configs/ch_PP-OCRv2_rec_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
rename to test_tipc/configs/ch_PP-OCRv3_rec/train_pact_infer_python.txt
index e22d8a564b..24469a91cf 100644
--- a/test_tipc/configs/ch_PP-OCRv2_rec_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_PP-OCRv3_rec/train_pact_infer_python.txt
@@ -1,20 +1,20 @@
===========================train_params===========================
-model_name:ch_PPOCRv2_rec_PACT
+model_name:ch_PP-OCRv3_rec_PACT
python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
-Global.auto_cast:amp
-Global.epoch_num:lite_train_lite_infer=3|whole_train_whole_infer=300
+Global.auto_cast:fp32
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=128
-Global.pretrained_model:null
+Global.pretrained_model:pretrain_models/ch_PP-OCRv3_rec_train/best_accuracy
train_model_name:latest
train_infer_img_dir:./inference/rec_inference
null:null
##
trainer:pact_train
norm_train:null
-pact_train:deploy/slim/quantization/quant.py -c test_tipc/configs/ch_PP-OCRv2_rec/ch_PP-OCRv2_rec_distillation.yml -o
+pact_train:deploy/slim/quantization/quant.py -c test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml -o
fpgm_train:null
distill_train:null
null:null
@@ -28,26 +28,26 @@ null:null
Global.save_inference_dir:./output/
Global.checkpoints:
norm_export:null
-quant_export:deploy/slim/quantization/export_model.py -c test_tipc/configs/ch_PP-OCRv2_rec/ch_PP-OCRv2_rec_distillation.yml -o
+quant_export:deploy/slim/quantization/export_model.py -c test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml -o
fpgm_export: null
distill_export:null
export1:null
export2:null
inference_dir:Student
-infer_model:./inference/ch_PP-OCRv2_rec_slim_quant_infer
+infer_model:./inference/ch_PP-OCRv3_rec_slim_infer
infer_export:null
infer_quant:True
-inference:tools/infer/predict_rec.py
+inference:tools/infer/predict_rec.py --rec_image_shape="3,48,320"
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:False|True
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
null:null
--benchmark:True
null:null
===========================infer_benchmark_params==========================
-random_infer_input:[{float32,[3,32,320]}]
+random_infer_input:[{float32,[3,48,320]}]
diff --git a/test_tipc/configs/ch_PP-OCRv3_rec/train_ptq_infer_python.txt b/test_tipc/configs/ch_PP-OCRv3_rec/train_ptq_infer_python.txt
new file mode 100644
index 0000000000..d1a8c7c001
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_rec/train_ptq_infer_python.txt
@@ -0,0 +1,21 @@
+===========================kl_quant_params===========================
+model_name:ch_PP-OCRv3_rec_KL
+python:python3.7
+Global.pretrained_model:
+Global.save_inference_dir:null
+infer_model:./inference/ch_PP-OCRv3_rec_infer/
+infer_export:deploy/slim/quantization/quant_kl.py -c test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml -o
+infer_quant:True
+inference:tools/infer/predict_rec.py --rec_image_shape="3,48,320"
+--use_gpu:False|True
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1|6
+--use_tensorrt:False
+--precision:int8
+--rec_model_dir:
+--image_dir:./inference/rec_inference
+null:null
+--benchmark:True
+null:null
+null:null
diff --git a/test_tipc/configs/ch_PP-OCRv3_rec_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_rec_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..f1a308fcc8
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_rec_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_PP-OCRv3_rec_KL
+use_opencv:True
+infer_model:./inference/ch_PP-OCRv3_rec_klquant_infer
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr --rec_img_h=48 --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference/
+null:null
+--benchmark:True
+--det:False
+--rec:True
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv3_rec_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_rec_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..fa6f04e48c
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_rec_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv3_rec_KL
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv3_det_klquant_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v3_kl_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v3_kl_client/
+--rec_dirname:./inference/ch_PP-OCRv3_rec_klquant_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v3_kl_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v3_kl_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv3_rec_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_rec_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..68586af0d9
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_rec_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv3_rec_KL
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:null
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:null
+--det_serving_client:null
+--rec_dirname:./inference/ch_PP-OCRv3_rec_klquant_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v3_kl_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v3_kl_client/
+serving_dir:./deploy/pdserving
+web_service:web_service_rec.py --config=config.yml --opt op.rec.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py --det=False
+--image_dir:../../inference/rec_inference
diff --git a/test_tipc/configs/ch_PP-OCRv3_rec_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_rec_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..8fc1132ca6
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_rec_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_PP-OCRv3_rec_PACT
+use_opencv:True
+infer_model:./inference/ch_PP-OCRv3_rec_pact_infer
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr --rec_img_h=48 --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference/
+null:null
+--benchmark:True
+--det:False
+--rec:True
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_PP-OCRv3_rec_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_rec_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..008df50d6d
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_rec_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv3_rec_PACT
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_PP-OCRv3_det_pact_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_v3_pact_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_v3_pact_client/
+--rec_dirname:./inference/ch_PP-OCRv3_rec_pact_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v3_pact_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v3_pact_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_PP-OCRv3_rec_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_rec_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..826f586f0c
--- /dev/null
+++ b/test_tipc/configs/ch_PP-OCRv3_rec_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_PP-OCRv3_rec_PACT
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:null
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:null
+--det_serving_client:null
+--rec_dirname:./inference/ch_PP-OCRv3_rec_pact_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_v3_pact_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_v3_pact_client/
+serving_dir:./deploy/pdserving
+web_service:web_service_rec.py --config=config.yml --opt op.rec.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py --det=False
+--image_dir:../../inference/rec_inference
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_PP-OCRv3_rec_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
similarity index 54%
rename from test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
rename to test_tipc/configs/ch_PP-OCRv3_rec_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index abed3cfba9..c93b307deb 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_PP-OCRv3_rec_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -1,20 +1,20 @@
===========================train_params===========================
-model_name:ch_ppocr_mobile_v2.0_rec_PACT
+model_name:ch_PP-OCRv3_rec_PACT
python:python3.7
-gpu_list:0
+gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:amp
-Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
Global.save_model_dir:./output/
-Train.loader.batch_size_per_card:lite_train_lite_infer=128|whole_train_whole_infer=128
-Global.checkpoints:null
+Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=128
+Global.pretrained_model:pretrain_models/ch_PP-OCRv3_rec_train/best_accuracy
train_model_name:latest
-train_infer_img_dir:./train_data/ic15_data/test/word_1.png
+train_infer_img_dir:./inference/rec_inference
null:null
##
trainer:pact_train
norm_train:null
-pact_train:deploy/slim/quantization/quant.py -c test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/rec_chinese_lite_train_v2.0.yml -o
+pact_train:deploy/slim/quantization/quant.py -c test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml -o
fpgm_train:null
distill_train:null
null:null
@@ -28,26 +28,26 @@ null:null
Global.save_inference_dir:./output/
Global.checkpoints:
norm_export:null
-quant_export:deploy/slim/quantization/export_model.py -c test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/rec_chinese_lite_train_v2.0.yml -o
-fpgm_export:null
+quant_export:deploy/slim/quantization/export_model.py -c test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml -o
+fpgm_export: null
distill_export:null
export1:null
export2:null
-inference_dir:null
-infer_model:./inference/ch_ppocr_mobile_v2.0_rec_slim_infer/
+inference_dir:Student
+infer_model:./inference/ch_PP-OCRv3_rec_slim_quant_infer
infer_export:null
-infer_quant:False
-inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt --rec_image_shape="3,32,100"
+infer_quant:True
+inference:tools/infer/predict_rec.py --rec_image_shape="3,48,320"
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
---rec_batch_num:1|6
---use_tensorrt:False|True
---precision:fp32|int8
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:6
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
---save_log_path:./test/output/
+null:null
--benchmark:True
null:null
===========================infer_benchmark_params==========================
-random_infer_input:[{float32,[3,32,320]}]
+random_infer_input:[{float32,[3,48,320]}]
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..b42ab9db36
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_ppocr_mobile_v2.0
+use_opencv:True
+infer_model:./inference/ch_ppocr_mobile_v2.0_det_infer/
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt --rec_img_h=32
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+--rec_model_dir:./inference/ch_ppocr_mobile_v2.0_rec_infer/
+--benchmark:True
+--det:True
+--rec:True
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
index 4a46f0cf09..becad991ea 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
@@ -6,10 +6,10 @@ infer_export:null
infer_quant:False
inference:tools/infer/predict_system.py
--use_gpu:False|True
---enable_mkldnn:False|True
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
+--use_tensorrt:False
--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
new file mode 100644
index 0000000000..17c2fbbae2
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
@@ -0,0 +1,17 @@
+===========================paddle2onnx_params===========================
+model_name:ch_ppocr_mobile_v2.0
+python:python3.7
+2onnx: paddle2onnx
+--det_model_dir:./inference/ch_ppocr_mobile_v2.0_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_save_file:./inference/det_mobile_onnx/model.onnx
+--rec_model_dir:./inference/ch_ppocr_mobile_v2.0_rec_infer/
+--rec_save_file:./inference/rec_mobile_onnx/model.onnx
+--opset_version:10
+--enable_onnx_checker:True
+inference:tools/infer/predict_system.py --rec_image_shape="3,32,320"
+--use_gpu:True|False
+--det_model_dir:
+--rec_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..d18e9f11fd
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_ppocr_mobile_v2.0
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_ppocr_mobile_v2.0_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_mobile_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_mobile_client/
+--rec_dirname:./inference/ch_ppocr_mobile_v2.0_rec_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_mobile_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_mobile_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..842c934017
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_ppocr_mobile_v2.0
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_ppocr_mobile_v2.0_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_mobile_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_mobile_client/
+--rec_dirname:./inference/ch_ppocr_mobile_v2.0_rec_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_mobile_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_mobile_client/
+serving_dir:./deploy/pdserving
+web_service:web_service.py --config=config.yml --opt op.det.concurrency="1" op.rec.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
index d0ae17ccb5..1d1c2ae283 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -1,15 +1,15 @@
===========================cpp_infer_params===========================
-model_name:ocr_det
+model_name:ch_ppocr_mobile_v2.0_det
use_opencv:True
infer_model:./inference/ch_ppocr_mobile_v2.0_det_infer/
infer_quant:False
inference:./deploy/cpp_infer/build/ppocr
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_infer_python_jetson.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_infer_python_jetson.txt
index 7d3f60bd42..24bb8746ab 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_infer_python_jetson.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_infer_python_jetson.txt
@@ -1,5 +1,5 @@
===========================infer_params===========================
-model_name:ocr_det
+model_name:ch_ppocr_mobile_v2.0_det
python:python
infer_model:./inference/ch_ppocr_mobile_v2.0_det_infer
infer_export:null
@@ -7,10 +7,10 @@ infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
--enable_mkldnn:False
---cpu_threads:1|6
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp16|fp32
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
index 160bcdbd88..00473d1062 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
@@ -1,14 +1,17 @@
===========================paddle2onnx_params===========================
-model_name:ocr_det_mobile
+model_name:ch_ppocr_mobile_v2.0_det
python:python3.7
2onnx: paddle2onnx
---model_dir:./inference/ch_ppocr_mobile_v2.0_det_infer/
+--det_model_dir:./inference/ch_ppocr_mobile_v2.0_det_infer/
--model_filename:inference.pdmodel
--params_filename:inference.pdiparams
---save_file:./inference/det_mobile_onnx/model.onnx
+--det_save_file:./inference/det_mobile_onnx/model.onnx
+--rec_model_dir:
+--rec_save_file:
--opset_version:10
--enable_onnx_checker:True
inference:tools/infer/predict_det.py
--use_gpu:True|False
--det_model_dir:
+--rec_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
index 2326c9d2a7..c9dd5ad920 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -1,18 +1,23 @@
===========================serving_params===========================
-model_name:ocr_det_mobile
-python:python3.7|cpp
+model_name:ch_ppocr_mobile_v2.0_det
+python:python3.7
trans_model:-m paddle_serving_client.convert
---dirname:./inference/ch_ppocr_mobile_v2.0_det_infer/
+--det_dirname:./inference/ch_ppocr_mobile_v2.0_det_infer/
--model_filename:inference.pdmodel
--params_filename:inference.pdiparams
---serving_server:./deploy/pdserving/ppocr_det_mobile_2.0_serving/
---serving_client:./deploy/pdserving/ppocr_det_mobile_2.0_client/
+--det_serving_server:./deploy/pdserving/ppocr_det_mobile_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_mobile_client/
+--rec_dirname:null
+--rec_serving_server:null
+--rec_serving_client:null
serving_dir:./deploy/pdserving
web_service:web_service_det.py --config=config.yml --opt op.det.concurrency="1"
-op.det.local_service_conf.devices:"0"|null
-op.det.local_service_conf.use_mkldnn:True|False
-op.det.local_service_conf.thread_num:1|6
-op.det.local_service_conf.use_trt:False|True
-op.det.local_service_conf.precision:fp32|fp16|int8
-pipline:pipeline_rpc_client.py|pipeline_http_client.py
---image_dir:../../doc/imgs
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_infer_python.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_infer_python.txt
index 269693a86e..789ed4d23d 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_infer_python.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_infer_python.txt
@@ -4,7 +4,7 @@ python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:null
-Global.epoch_num:lite_train_lite_infer=100|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=100|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_linux_gpu_fleet_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_linux_gpu_fleet_amp_infer_python_linux_gpu_cpu.txt
deleted file mode 100644
index bfb71b7810..0000000000
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_linux_gpu_fleet_amp_infer_python_linux_gpu_cpu.txt
+++ /dev/null
@@ -1,51 +0,0 @@
-===========================train_params===========================
-model_name:ocr_det
-python:python3.7
-gpu_list:xx.xx.xx.xx,yy.yy.yy.yy;0,1
-Global.use_gpu:True
-Global.auto_cast:fp32|amp
-Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=300
-Global.save_model_dir:./output/
-Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
-Global.pretrained_model:null
-train_model_name:latest
-train_infer_img_dir:./train_data/icdar2015/text_localization/ch4_test_images/
-null:null
-##
-trainer:norm_train|pact_train|fpgm_train
-norm_train:tools/train.py -c test_tipc/configs/ppocr_det_mobile/det_mv3_db.yml -o Global.pretrained_model=./pretrain_models/MobileNetV3_large_x0_5_pretrained
-pact_train:deploy/slim/quantization/quant.py -c test_tipc/configs/ppocr_det_mobile/det_mv3_db.yml -o
-fpgm_train:deploy/slim/prune/sensitivity_anal.py -c test_tipc/configs/ppocr_det_mobile/det_mv3_db.yml -o Global.pretrained_model=./pretrain_models/det_mv3_db_v2.0_train/best_accuracy
-distill_train:null
-null:null
-null:null
-##
-===========================eval_params===========================
-eval:null
-null:null
-##
-===========================infer_params===========================
-Global.save_inference_dir:./output/
-Global.pretrained_model:
-norm_export:tools/export_model.py -c test_tipc/configs/ppocr_det_mobile/det_mv3_db.yml -o
-quant_export:deploy/slim/quantization/export_model.py -c test_tipc/configs/ppocr_det_mobile/det_mv3_db.yml -o
-fpgm_export:deploy/slim/prune/export_prune_model.py -c test_tipc/configs/ppocr_det_mobile/det_mv3_db.yml -o
-distill_export:null
-export1:null
-export2:null
-inference_dir:null
-train_model:./inference/ch_ppocr_mobile_v2.0_det_train/best_accuracy
-infer_export:tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o
-infer_quant:False
-inference:tools/infer/predict_det.py
---use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
---rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
---det_model_dir:
---image_dir:./inference/ch_det_data_50/all-sum-510/
-null:null
---benchmark:True
-null:null
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..5271f78bb7
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:ch_ppocr_mobile_v2.0_det
+python:python3.7
+gpu_list:192.168.0.1,192.168.0.2;0,1
+Global.use_gpu:True
+Global.auto_cast:null
+Global.epoch_num:lite_train_lite_infer=100|whole_train_whole_infer=50
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./train_data/icdar2015/text_localization/ch4_test_images/
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model=./pretrain_models/MobileNetV3_large_x0_5_pretrained
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:null
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o
+quant_export:null
+fpgm_export:null
+distill_export:null
+export1:null
+export2:null
+inference_dir:null
+train_model:./inference/ch_ppocr_mobile_v2.0_det_train/best_accuracy
+infer_export:tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o
+infer_quant:False
+inference:tools/infer/predict_det.py
+--use_gpu:False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,640,640]}];[{float32,[3,960,960]}]
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 593e7ec7ed..6b3352f741 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -4,7 +4,7 @@ python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:amp
-Global.epoch_num:lite_train_lite_infer=100|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=100|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_mac_cpu_normal_normal_infer_python_mac_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_mac_cpu_normal_normal_infer_python_mac_cpu.txt
index 014dad5fc9..3f321a1903 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_mac_cpu_normal_normal_infer_python_mac_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_mac_cpu_normal_normal_infer_python_mac_cpu.txt
@@ -4,7 +4,7 @@ python:python
gpu_list:-1
Global.use_gpu:False
Global.auto_cast:null
-Global.epoch_num:lite_train_lite_infer=5|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=5|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
Global.pretrained_model:null
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/train_infer_python.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_pact_infer_python.txt
similarity index 89%
rename from test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/train_infer_python.txt
rename to test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_pact_infer_python.txt
index 9d2855d824..04c8d0e194 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/train_infer_python.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_pact_infer_python.txt
@@ -4,7 +4,7 @@ python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:null
-Global.epoch_num:lite_train_lite_infer=20|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=2|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:null
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_ptq_infer_python.txt
similarity index 89%
rename from test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
rename to test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_ptq_infer_python.txt
index 1039dcad06..2bdec84883 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_ptq_infer_python.txt
@@ -8,10 +8,10 @@ infer_export:deploy/slim/quantization/quant_kl.py -c configs/det/ch_ppocr_v2.0/c
infer_quant:True
inference:tools/infer/predict_det.py
--use_gpu:False|True
---enable_mkldnn:True
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
+--use_tensorrt:False
--precision:int8
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_windows_gpu_normal_normal_infer_python_windows_cpu_gpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_windows_gpu_normal_normal_infer_python_windows_cpu_gpu.txt
index 6a63b39d97..a3f6933a64 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_windows_gpu_normal_normal_infer_python_windows_cpu_gpu.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_windows_gpu_normal_normal_infer_python_windows_cpu_gpu.txt
@@ -4,7 +4,7 @@ python:python
gpu_list:0
Global.use_gpu:True
Global.auto_cast:fp32|amp
-Global.epoch_num:lite_train_lite_infer=5|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=5|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
Global.pretrained_model:null
@@ -39,10 +39,10 @@ infer_export:tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
+--enable_mkldnn:False
--cpu_threads:1|6
--rec_batch_num:1
---use_tensorrt:False|True
+--use_tensorrt:False
--precision:fp32|fp16|int8
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_FPGM/train_infer_python.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_FPGM/train_infer_python.txt
index 47ccf2e69e..dae3f8053a 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_FPGM/train_infer_python.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_FPGM/train_infer_python.txt
@@ -4,7 +4,7 @@ python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:null
-Global.epoch_num:lite_train_lite_infer=5|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=5|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:null
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_FPGM/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_FPGM/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 5a95f02685..150a8a0315 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_FPGM/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_FPGM/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -4,7 +4,7 @@ python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:amp
-Global.epoch_num:lite_train_lite_infer=5|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=5|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:null
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..eb2fd0a001
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_ppocr_mobile_v2.0_det_KL
+use_opencv:True
+infer_model:./inference/ch_ppocr_mobile_v2.0_det_klquant_infer
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+--det:True
+--rec:False
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..d9de1cc19a
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_ppocr_mobile_v2.0_det_KL
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_ppocr_mobile_v2.0_det_klquant_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_mobile_kl_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_mobile_kl_client/
+--rec_dirname:./inference/ch_ppocr_mobile_v2.0_rec_klquant_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_mobile_kl_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_mobile_kl_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..049ec78458
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_ppocr_mobile_v2.0_det_KL
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_ppocr_mobile_v2.0_det_klquant_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_mobile_kl_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_mobile_kl_client/
+--rec_dirname:null
+--rec_serving_server:null
+--rec_serving_client:null
+serving_dir:./deploy/pdserving
+web_service:web_service_det.py --config=config.yml --opt op.det.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..17723f41ab
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_ppocr_mobile_v2.0_det_PACT
+use_opencv:True
+infer_model:./inference/ch_ppocr_mobile_v2.0_det_pact_infer
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+--det:True
+--rec:False
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..1a49a10f9b
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_ppocr_mobile_v2.0_det_PACT
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_ppocr_mobile_v2.0_det_pact_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_mobile_pact_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_mobile_pact_client/
+--rec_dirname:./inference/ch_ppocr_mobile_v2.0_rec_pact_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_mobile_pact_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_mobile_pact_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..909d738919
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_ppocr_mobile_v2.0_det_PACT
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_ppocr_mobile_v2.0_det_pact_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_mobile_pact_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_mobile_pact_client/
+--rec_dirname:null
+--rec_serving_server:null
+--rec_serving_client:null
+serving_dir:./deploy/pdserving
+web_service:web_service_det.py --config=config.yml --opt op.det.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..480fb16cdd
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_ppocr_mobile_v2.0_rec
+use_opencv:True
+infer_model:./inference/ch_ppocr_mobile_v2.0_rec_infer/
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt --rec_img_h=32
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference/
+null:null
+--benchmark:True
+--det:False
+--rec:True
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
index f29b303879..5bab0c9e4c 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
@@ -1,14 +1,17 @@
===========================paddle2onnx_params===========================
-model_name:ocr_rec_mobile
+model_name:ch_ppocr_mobile_v2.0_rec
python:python3.7
2onnx: paddle2onnx
---model_dir:./inference/ch_ppocr_mobile_v2.0_rec_infer/
+--det_model_dir:
--model_filename:inference.pdmodel
--params_filename:inference.pdiparams
---save_file:./inference/rec_mobile_onnx/model.onnx
+--det_save_file:
+--rec_model_dir:./inference/ch_ppocr_mobile_v2.0_rec_infer/
+--rec_save_file:./inference/rec_mobile_onnx/model.onnx
--opset_version:10
--enable_onnx_checker:True
-inference:tools/infer/predict_rec.py
+inference:tools/infer/predict_rec.py --rec_image_shape="3,32,320"
--use_gpu:True|False
+--det_model_dir:
--rec_model_dir:
---image_dir:./inference/rec_inference
\ No newline at end of file
+--image_dir:./inference/rec_inference/
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
index f890eff469..c0c5291cc4 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -1,18 +1,23 @@
===========================serving_params===========================
-model_name:ocr_rec_mobile
-python:python3.7|cpp
+model_name:ch_ppocr_mobile_v2.0_rec
+python:python3.7
trans_model:-m paddle_serving_client.convert
---dirname:./inference/ch_ppocr_mobile_v2.0_rec_infer/
+--det_dirname:null
--model_filename:inference.pdmodel
--params_filename:inference.pdiparams
---serving_server:./deploy/pdserving/ppocr_rec_mobile_2.0_serving/
---serving_client:./deploy/pdserving/ppocr_rec_mobile_2.0_client/
+--det_serving_server:null
+--det_serving_client:null
+--rec_dirname:./inference/ch_ppocr_mobile_v2.0_rec_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_mobile_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_mobile_client/
serving_dir:./deploy/pdserving
-web_service:web_service_rec.py --config=config.yml --opt op.rec.concurrency=1
-op.rec.local_service_conf.devices:"0"|null
-op.rec.local_service_conf.use_mkldnn:True|False
-op.rec.local_service_conf.thread_num:1|6
-op.rec.local_service_conf.use_trt:False|True
-op.rec.local_service_conf.precision:fp32|fp16|int8
-pipline:pipeline_rpc_client.py|pipeline_http_client.py
---image_dir:../../doc/imgs_words_en
\ No newline at end of file
+web_service:web_service_rec.py --config=config.yml --opt op.rec.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py --det=False
+--image_dir:../../inference/rec_inference
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_infer_python.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_infer_python.txt
index 5086f80d7b..f02b93926c 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_infer_python.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_infer_python.txt
@@ -4,7 +4,7 @@ python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:null
-Global.epoch_num:lite_train_lite_infer=2|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=2|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=128|whole_train_whole_infer=128
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c configs/rec/rec_icdar15_train.yml -o
infer_quant:False
inference:tools/infer/predict_rec.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..631118c0a9
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:ch_ppocr_mobile_v2.0_rec
+python:python3.7
+gpu_list:192.168.0.1,192.168.0.2;0,1
+Global.use_gpu:True
+Global.auto_cast:null
+Global.epoch_num:lite_train_lite_infer=2|whole_train_whole_infer=50
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=128|whole_train_whole_infer=128
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./inference/rec_inference
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c configs/rec/rec_icdar15_train.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:tools/eval.py -c configs/rec/rec_icdar15_train.yml -o
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c configs/rec/rec_icdar15_train.yml -o
+quant_export:null
+fpgm_export:null
+distill_export:null
+export1:null
+export2:null
+##
+train_model:./inference/ch_ppocr_mobile_v2.0_rec_train/best_accuracy
+infer_export:tools/export_model.py -c configs/rec/rec_icdar15_train.yml -o
+infer_quant:False
+inference:tools/infer/predict_rec.py
+--use_gpu:False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1|6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference
+--save_log_path:./test/output/
+--benchmark:True
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,32,100]}]
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 30fb939bff..bd9c4a8df2 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -4,7 +4,7 @@ python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:amp
-Global.epoch_num:lite_train_lite_infer=2|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=2|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=128|whole_train_whole_infer=128
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c configs/rec/rec_icdar15_train.yml -o
infer_quant:False
inference:tools/infer/predict_rec.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/train_infer_python.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_pact_infer_python.txt
similarity index 94%
rename from test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/train_infer_python.txt
rename to test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_pact_infer_python.txt
index 94909ec340..77472fbdfb 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/train_infer_python.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_pact_infer_python.txt
@@ -4,7 +4,7 @@ python:python3.7
gpu_list:0
Global.use_gpu:True|True
Global.auto_cast:null
-Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=128|whole_train_whole_infer=128
Global.checkpoints:null
@@ -39,11 +39,11 @@ infer_export:null
infer_quant:False
inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt --rec_image_shape="3,32,100"
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:False|True
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_ptq_infer_python.txt
similarity index 81%
rename from test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
rename to test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_ptq_infer_python.txt
index 4b77994f3f..f63fe4c2bb 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_ptq_infer_python.txt
@@ -6,12 +6,12 @@ Global.save_inference_dir:null
infer_model:./inference/ch_ppocr_mobile_v2.0_rec_infer/
infer_export:deploy/slim/quantization/quant_kl.py -c test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/rec_chinese_lite_train_v2.0.yml -o
infer_quant:True
-inference:tools/infer/predict_rec.py
+inference:tools/infer/predict_rec.py --rec_image_shape="3,32,320"
--use_gpu:False|True
---enable_mkldnn:True
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
+--use_tensorrt:False
--precision:int8
--rec_model_dir:
--image_dir:./inference/rec_inference
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_FPGM/train_infer_python.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_FPGM/train_infer_python.txt
index 77494ac347..89daceeb5f 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_FPGM/train_infer_python.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_FPGM/train_infer_python.txt
@@ -4,7 +4,7 @@ python:python3.7
gpu_list:0
Global.use_gpu:True|True
Global.auto_cast:null
-Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=128|whole_train_whole_infer=128
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:null
infer_quant:False
inference:tools/infer/predict_rec.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
null:null
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_FPGM/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_FPGM/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index fda9cf4dde..7abc3e9340 100644
--- a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_FPGM/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_FPGM/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -4,7 +4,7 @@ python:python3.7
gpu_list:0
Global.use_gpu:True|True
Global.auto_cast:amp
-Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=128|whole_train_whole_infer=128
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:null
infer_quant:False
inference:tools/infer/predict_rec.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
null:null
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..adf06257a7
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_ppocr_mobile_v2.0_rec_KL
+use_opencv:True
+infer_model:./inference/ch_ppocr_mobile_v2.0_rec_klquant_infer
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt --rec_img_h=32
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference/
+null:null
+--benchmark:True
+--det:False
+--rec:True
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..ab518de55a
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_ppocr_mobile_v2.0_rec_KL
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_ppocr_mobile_v2.0_det_klquant_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_mobile_kl_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_mobile_kl_client/
+--rec_dirname:./inference/ch_ppocr_mobile_v2.0_rec_klquant_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_mobile_kl_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_mobile_kl_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..948e3dceb3
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_ppocr_mobile_v2.0_rec_KL
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:null
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:null
+--det_serving_client:null
+--rec_dirname:./inference/ch_ppocr_mobile_v2.0_rec_klquant_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_mobile_kl_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_mobile_kl_client/
+serving_dir:./deploy/pdserving
+web_service:web_service_rec.py --config=config.yml --opt op.rec.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py --det=False
+--image_dir:../../inference/rec_inference
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..ba2df90f75
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_ppocr_mobile_v2.0_rec_PACT
+use_opencv:True
+infer_model:./inference/ch_ppocr_mobile_v2.0_rec_pact_infer
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt --rec_img_h=32
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference/
+null:null
+--benchmark:True
+--det:False
+--rec:True
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..229f70cf35
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_ppocr_mobile_v2.0_rec_PACT
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_ppocr_mobile_v2.0_det_pact_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_mobile_pact_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_mobile_pact_client/
+--rec_dirname:./inference/ch_ppocr_mobile_v2.0_rec_pact_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_mobile_pact_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_mobile_pact_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..f123f36543
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_ppocr_mobile_v2.0_rec_PACT
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:null
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:null
+--det_serving_client:null
+--rec_dirname:./inference/ch_ppocr_mobile_v2.0_rec_pact_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_mobile_pact_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_mobile_pact_client/
+serving_dir:./deploy/pdserving
+web_service:web_service_rec.py --config=config.yml --opt op.rec.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py --det=False
+--image_dir:../../inference/rec_inference
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..7c980b2bae
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_ppocr_server_v2.0
+use_opencv:True
+infer_model:./inference/ch_ppocr_server_v2.0_det_infer/
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt --rec_img_h=32
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+--rec_model_dir:./inference/ch_ppocr_server_v2.0_rec_infer/
+--benchmark:True
+--det:True
+--rec:True
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
index 92d7031e88..b20596f7a1 100644
--- a/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
@@ -6,8 +6,8 @@ infer_export:null
infer_quant:True
inference:tools/infer/predict_system.py
--use_gpu:False|True
---enable_mkldnn:False|True
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
--use_tensorrt:False
--precision:fp32
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
new file mode 100644
index 0000000000..e478896a54
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
@@ -0,0 +1,17 @@
+===========================paddle2onnx_params===========================
+model_name:ch_ppocr_server_v2.0
+python:python3.7
+2onnx: paddle2onnx
+--det_model_dir:./inference/ch_ppocr_server_v2.0_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_save_file:./inference/det_server_onnx/model.onnx
+--rec_model_dir:./inference/ch_ppocr_server_v2.0_rec_infer/
+--rec_save_file:./inference/rec_server_onnx/model.onnx
+--opset_version:10
+--enable_onnx_checker:True
+inference:tools/infer/predict_system.py --rec_image_shape="3,32,320"
+--use_gpu:True|False
+--det_model_dir:
+--rec_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/00008790.jpg
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..bbfec44dba
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,19 @@
+===========================serving_params===========================
+model_name:ch_ppocr_server_v2.0
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_ppocr_server_v2.0_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_server_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_server_client/
+--rec_dirname:./inference/ch_ppocr_server_v2.0_rec_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_server_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_server_client/
+serving_dir:./deploy/pdserving
+web_service:-m paddle_serving_server.serve
+--op:GeneralDetectionOp GeneralInferOp
+--port:8181
+--gpu_id:"0"|null
+cpp_client:ocr_cpp_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..8853e709d4
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -0,0 +1,23 @@
+===========================serving_params===========================
+model_name:ch_ppocr_server_v2.0
+python:python3.7
+trans_model:-m paddle_serving_client.convert
+--det_dirname:./inference/ch_ppocr_server_v2.0_det_infer/
+--model_filename:inference.pdmodel
+--params_filename:inference.pdiparams
+--det_serving_server:./deploy/pdserving/ppocr_det_server_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_server_client/
+--rec_dirname:./inference/ch_ppocr_server_v2.0_rec_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_server_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_server_client/
+serving_dir:./deploy/pdserving
+web_service:web_service.py --config=config.yml --opt op.det.concurrency="1" op.rec.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..69ae939e2b
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_server_v2.0_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_ppocr_server_v2.0_det
+use_opencv:True
+infer_model:./inference/ch_ppocr_server_v2.0_det_infer/
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+null:null
+--benchmark:True
+--det:True
+--rec:False
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
index 40fdc11241..c8bebf54f2 100644
--- a/test_tipc/configs/ch_ppocr_server_v2.0_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_server_v2.0_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
@@ -1,14 +1,17 @@
===========================paddle2onnx_params===========================
-model_name:ocr_det_server
+model_name:ch_ppocr_server_v2.0_det
python:python3.7
2onnx: paddle2onnx
---model_dir:./inference/ch_ppocr_server_v2.0_det_infer/
+--det_model_dir:./inference/ch_ppocr_server_v2.0_det_infer/
--model_filename:inference.pdmodel
--params_filename:inference.pdiparams
---save_file:./inference/det_server_onnx/model.onnx
+--det_save_file:./inference/det_server_onnx/model.onnx
+--rec_model_dir:
+--rec_save_file:
--opset_version:10
--enable_onnx_checker:True
inference:tools/infer/predict_det.py
--use_gpu:True|False
--det_model_dir:
---image_dir:./inference/det_inference
\ No newline at end of file
+--rec_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/00008790.jpg
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
index ec54646046..018dd1a227 100644
--- a/test_tipc/configs/ch_ppocr_server_v2.0_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_server_v2.0_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -1,18 +1,23 @@
===========================serving_params===========================
-model_name:ocr_det_server
-python:python3.7|cpp
+model_name:ch_ppocr_server_v2.0_det
+python:python3.7
trans_model:-m paddle_serving_client.convert
---dirname:./inference/ch_ppocr_server_v2.0_det_infer/
+--det_dirname:./inference/ch_ppocr_server_v2.0_det_infer/
--model_filename:inference.pdmodel
--params_filename:inference.pdiparams
---serving_server:./deploy/pdserving/ppocr_det_server_2.0_serving/
---serving_client:./deploy/pdserving/ppocr_det_server_2.0_client/
+--det_serving_server:./deploy/pdserving/ppocr_det_server_serving/
+--det_serving_client:./deploy/pdserving/ppocr_det_server_client/
+--rec_dirname:null
+--rec_serving_server:null
+--rec_serving_client:null
serving_dir:./deploy/pdserving
-web_service:web_service_det.py --config=config.yml --opt op.det.concurrency=1
-op.det.local_service_conf.devices:"0"|null
-op.det.local_service_conf.use_mkldnn:True|False
-op.det.local_service_conf.thread_num:1|6
-op.det.local_service_conf.use_trt:False|True
-op.det.local_service_conf.precision:fp32|fp16|int8
-pipline:pipeline_rpc_client.py|pipeline_http_client.py
---image_dir:../../doc/imgs
\ No newline at end of file
+web_service:web_service_det.py --config=config.yml --opt op.det.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py
+--image_dir:../../doc/imgs/1.jpg
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0_det/train_infer_python.txt b/test_tipc/configs/ch_ppocr_server_v2.0_det/train_infer_python.txt
index 52489fe529..c16ca15002 100644
--- a/test_tipc/configs/ch_ppocr_server_v2.0_det/train_infer_python.txt
+++ b/test_tipc/configs/ch_ppocr_server_v2.0_det/train_infer_python.txt
@@ -4,7 +4,7 @@ python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:null
-Global.epoch_num:lite_train_lite_infer=2|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=2|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_lite_infer=4
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_res18_db_
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
--save_log_path:null
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..12388d9677
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_server_v2.0_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:ch_ppocr_server_v2.0_det
+python:python3.7
+gpu_list:192.168.0.1,192.168.0.2;0,1
+Global.use_gpu:True
+Global.auto_cast:null
+Global.epoch_num:lite_train_lite_infer=2|whole_train_whole_infer=50
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_lite_infer=4
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./train_data/icdar2015/text_localization/ch4_test_images/
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c test_tipc/configs/ch_ppocr_server_v2.0_det/det_r50_vd_db.yml -o
+quant_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:tools/eval.py -c test_tipc/configs/ch_ppocr_server_v2.0_det/det_r50_vd_db.yml -o
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c test_tipc/configs/ch_ppocr_server_v2.0_det/det_r50_vd_db.yml -o
+quant_export:null
+fpgm_export:null
+distill_export:null
+export1:null
+export2:null
+##
+train_model:./inference/ch_ppocr_server_v2.0_det_train/best_accuracy
+infer_export:tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_res18_db_v2.0.yml -o
+infer_quant:False
+inference:tools/infer/predict_det.py
+--use_gpu:False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--det_model_dir:
+--image_dir:./inference/ch_det_data_50/all-sum-510/
+--save_log_path:null
+--benchmark:True
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,640,640]}];[{float32,[3,960,960]}]
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 3e3764e8c6..93ed14cb60 100644
--- a/test_tipc/configs/ch_ppocr_server_v2.0_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_server_v2.0_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -4,7 +4,7 @@ python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:amp
-Global.epoch_num:lite_train_lite_infer=2|whole_train_whole_infer=300
+Global.epoch_num:lite_train_lite_infer=2|whole_train_whole_infer=50
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_lite_infer=4
Global.pretrained_model:null
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_res18_db_
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
--save_log_path:null
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..cbec272cce
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_server_v2.0_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
@@ -0,0 +1,20 @@
+===========================cpp_infer_params===========================
+model_name:ch_ppocr_server_v2.0_rec
+use_opencv:True
+infer_model:./inference/ch_ppocr_server_v2.0_rec_infer/
+infer_quant:False
+inference:./deploy/cpp_infer/build/ppocr --rec_char_dict_path=./ppocr/utils/ppocr_keys_v1.txt --rec_img_h=32
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference/
+null:null
+--benchmark:True
+--det:False
+--rec:True
+--cls:False
+--use_angle_cls:False
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
index 05542332e9..462f6090d9 100644
--- a/test_tipc/configs/ch_ppocr_server_v2.0_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_server_v2.0_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
@@ -1,14 +1,17 @@
===========================paddle2onnx_params===========================
-model_name:ocr_rec_server
+model_name:ch_ppocr_server_v2.0_rec
python:python3.7
2onnx: paddle2onnx
---model_dir:./inference/ch_ppocr_server_v2.0_rec_infer/
+--det_model_dir:
--model_filename:inference.pdmodel
--params_filename:inference.pdiparams
---save_file:./inference/rec_server_onnx/model.onnx
+--det_save_file:
+--rec_model_dir:./inference/ch_ppocr_server_v2.0_rec_infer/
+--rec_save_file:./inference/rec_server_onnx/model.onnx
--opset_version:10
--enable_onnx_checker:True
-inference:tools/infer/predict_rec.py
+inference:tools/infer/predict_rec.py --rec_image_shape="3,32,320"
--use_gpu:True|False
+--det_model_dir:
--rec_model_dir:
---image_dir:./inference/rec_inference
\ No newline at end of file
+--image_dir:./inference/rec_inference/
\ No newline at end of file
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
index d72abc6054..7f456320b6 100644
--- a/test_tipc/configs/ch_ppocr_server_v2.0_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_server_v2.0_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
@@ -1,18 +1,23 @@
===========================serving_params===========================
-model_name:ocr_rec_server
-python:python3.7|cpp
+model_name:ch_ppocr_server_v2.0_rec
+python:python3.7
trans_model:-m paddle_serving_client.convert
---dirname:./inference/ch_ppocr_server_v2.0_rec_infer/
+--det_dirname:null
--model_filename:inference.pdmodel
--params_filename:inference.pdiparams
---serving_server:./deploy/pdserving/ppocr_rec_server_2.0_serving/
---serving_client:./deploy/pdserving/ppocr_rec_server_2.0_client/
+--det_serving_server:null
+--det_serving_client:null
+--rec_dirname:./inference/ch_ppocr_server_v2.0_rec_infer/
+--rec_serving_server:./deploy/pdserving/ppocr_rec_server_serving/
+--rec_serving_client:./deploy/pdserving/ppocr_rec_server_client/
serving_dir:./deploy/pdserving
-web_service:web_service_rec.py --config=config.yml --opt op.rec.concurrency=1
-op.rec.local_service_conf.devices:"0"|null
-op.rec.local_service_conf.use_mkldnn:True|False
-op.rec.local_service_conf.thread_num:1|6
-op.rec.local_service_conf.use_trt:False|True
-op.rec.local_service_conf.precision:fp32|fp16|int8
-pipline:pipeline_rpc_client.py|pipeline_http_client.py
---image_dir:../../doc/imgs_words_en
\ No newline at end of file
+web_service:web_service_rec.py --config=config.yml --opt op.rec.concurrency="1"
+op.det.local_service_conf.devices:gpu|null
+op.det.local_service_conf.use_mkldnn:False
+op.det.local_service_conf.thread_num:6
+op.det.local_service_conf.use_trt:False
+op.det.local_service_conf.precision:fp32
+op.det.local_service_conf.model_config:
+op.rec.local_service_conf.model_config:
+pipline:pipeline_http_client.py --det=False
+--image_dir:../../inference/rec_inference
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0_rec/train_infer_python.txt b/test_tipc/configs/ch_ppocr_server_v2.0_rec/train_infer_python.txt
index 78a046c503..64c0cf455c 100644
--- a/test_tipc/configs/ch_ppocr_server_v2.0_rec/train_infer_python.txt
+++ b/test_tipc/configs/ch_ppocr_server_v2.0_rec/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/ch_ppocr_server_v2.0_rec
infer_quant:False
inference:tools/infer/predict_rec.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..9884ab247b
--- /dev/null
+++ b/test_tipc/configs/ch_ppocr_server_v2.0_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:ch_ppocr_server_v2.0_rec
+python:python3.7
+gpu_list:192.168.0.1,192.168.0.2;0,1
+Global.use_gpu:True
+Global.auto_cast:null
+Global.epoch_num:lite_train_lite_infer=5|whole_train_whole_infer=100
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=128|whole_train_whole_infer=128
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./inference/rec_inference
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c test_tipc/configs/ch_ppocr_server_v2.0_rec/rec_icdar15_train.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:tools/eval.py -c test_tipc/configs/ch_ppocr_server_v2.0_rec/rec_icdar15_train.yml -o
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c test_tipc/configs/ch_ppocr_server_v2.0_rec/rec_icdar15_train.yml -o
+quant_export:null
+fpgm_export:null
+distill_export:null
+export1:null
+export2:null
+##
+train_model:./inference/ch_ppocr_server_v2.0_rec_train/best_accuracy
+infer_export:tools/export_model.py -c test_tipc/configs/ch_ppocr_server_v2.0_rec/rec_icdar15_train.yml -o
+infer_quant:False
+inference:tools/infer/predict_rec.py
+--use_gpu:False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1|6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference
+--save_log_path:./test/output/
+--benchmark:True
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,32,100]}]
diff --git a/test_tipc/configs/ch_ppocr_server_v2.0_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ch_ppocr_server_v2.0_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 78c15047fb..63ddaa4a8b 100644
--- a/test_tipc/configs/ch_ppocr_server_v2.0_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ch_ppocr_server_v2.0_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/ch_ppocr_server_v2.0_rec
infer_quant:False
inference:tools/infer/predict_rec.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/det_mv3_db_v2_0/train_infer_python.txt b/test_tipc/configs/det_mv3_db_v2_0/train_infer_python.txt
index fab8f50d54..ab3aa59b60 100644
--- a/test_tipc/configs/det_mv3_db_v2_0/train_infer_python.txt
+++ b/test_tipc/configs/det_mv3_db_v2_0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c configs/det/det_mv3_db.yml -o
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/det_mv3_east_v2.0/train_infer_python.txt b/test_tipc/configs/det_mv3_east_v2.0/train_infer_python.txt
index 5634297973..1ec1597a4d 100644
--- a/test_tipc/configs/det_mv3_east_v2.0/train_infer_python.txt
+++ b/test_tipc/configs/det_mv3_east_v2.0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/det_mv3_east_v2.0/det_mv
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
--use_tensorrt:False
---precision:fp32|fp16|int8
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
--save_log_path:null
diff --git a/test_tipc/configs/det_mv3_pse_v2.0/train_infer_python.txt b/test_tipc/configs/det_mv3_pse_v2.0/train_infer_python.txt
index 661adc4a32..daeec69f84 100644
--- a/test_tipc/configs/det_mv3_pse_v2.0/train_infer_python.txt
+++ b/test_tipc/configs/det_mv3_pse_v2.0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/det_mv3_pse_v2.0/det_mv3
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
--use_tensorrt:False
---precision:fp32|fp16
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
--save_log_path:null
diff --git a/test_tipc/configs/det_r18_vd_db_v2_0/train_infer_python.txt b/test_tipc/configs/det_r18_vd_db_v2_0/train_infer_python.txt
index 77023ef2c1..33e4dbf233 100644
--- a/test_tipc/configs/det_r18_vd_db_v2_0/train_infer_python.txt
+++ b/test_tipc/configs/det_r18_vd_db_v2_0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:null
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
--save_log_path:null
diff --git a/test_tipc/configs/det_r50_db_v2.0/train_infer_python.txt b/test_tipc/configs/det_r50_db_v2.0/train_infer_python.txt
index 3fd875711e..151f2769cc 100644
--- a/test_tipc/configs/det_r50_db_v2.0/train_infer_python.txt
+++ b/test_tipc/configs/det_r50_db_v2.0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c configs/det/det_r50_vd_db.yml -o
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
---use_tensorrt:False|True
---precision:fp32|fp16|int8
+--use_tensorrt:False
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
--save_log_path:null
diff --git a/test_tipc/configs/det_r50_vd_east_v2_0/train_infer_python.txt b/test_tipc/configs/det_r50_vd_east_v2_0/train_infer_python.txt
index c1748c5d2f..8477a4fa74 100644
--- a/test_tipc/configs/det_r50_vd_east_v2_0/train_infer_python.txt
+++ b/test_tipc/configs/det_r50_vd_east_v2_0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/det_r50_vd_east_v2_0/det
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
--use_tensorrt:False
---precision:fp32|fp16|int8
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
--save_log_path:null
diff --git a/test_tipc/configs/det_r50_vd_pse_v2_0/train_infer_python.txt b/test_tipc/configs/det_r50_vd_pse_v2_0/train_infer_python.txt
index 55ebcd3547..62da89fe1c 100644
--- a/test_tipc/configs/det_r50_vd_pse_v2_0/train_infer_python.txt
+++ b/test_tipc/configs/det_r50_vd_pse_v2_0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/det_r50_vd_pse_v2_0/det_
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
--use_tensorrt:False
---precision:fp32|fp16|int8
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
--save_log_path:null
diff --git a/test_tipc/configs/det_r50_vd_sast_icdar15_v2.0/train_infer_python.txt b/test_tipc/configs/det_r50_vd_sast_icdar15_v2.0/train_infer_python.txt
index 16f37ace6f..b70ef46b4a 100644
--- a/test_tipc/configs/det_r50_vd_sast_icdar15_v2.0/train_infer_python.txt
+++ b/test_tipc/configs/det_r50_vd_sast_icdar15_v2.0/train_infer_python.txt
@@ -4,16 +4,16 @@ python:python3.7
gpu_list:0|0,1
Global.use_gpu:True|True
Global.auto_cast:null
-Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=5000
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=500
Global.save_model_dir:./output/
Train.loader.batch_size_per_card:lite_train_lite_infer=2|whole_train_whole_infer=4
-Global.pretrained_model:null
+Global.pretrained_model:./pretrain_models/det_r50_vd_sast_icdar15_v2.0_train/best_accuracy
train_model_name:latest
train_infer_img_dir:./train_data/icdar2015/text_localization/ch4_test_images/
null:null
##
trainer:norm_train
-norm_train:tools/train.py -c test_tipc/configs/det_r50_vd_sast_icdar15_v2.0/det_r50_vd_sast_icdar2015.yml -o Global.pretrained_model=./pretrain_models/ResNet50_vd_ssld_pretrained
+norm_train:tools/train.py -c test_tipc/configs/det_r50_vd_sast_icdar15_v2.0/det_r50_vd_sast_icdar2015.yml -o
pact_train:null
fpgm_train:null
distill_train:null
@@ -39,13 +39,13 @@ infer_export:tools/export_model.py -c test_tipc/configs/det_r50_vd_sast_icdar15_
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
--use_tensorrt:False
---precision:fp32|int8
+--precision:fp32
--det_model_dir:
---image_dir:./inference/ch_det_data_50/all-sum-510/
+--image_dir:./inference/ch_det_data_50/all-sum-510/00008790.jpg
null:null
--benchmark:True
--det_algorithm:SAST
diff --git a/test_tipc/configs/det_r50_vd_sast_totaltext_v2.0/train_infer_python.txt b/test_tipc/configs/det_r50_vd_sast_totaltext_v2.0/train_infer_python.txt
index 5e4c5666b7..7be5af7dde 100644
--- a/test_tipc/configs/det_r50_vd_sast_totaltext_v2.0/train_infer_python.txt
+++ b/test_tipc/configs/det_r50_vd_sast_totaltext_v2.0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/det_r50_vd_sast_totaltex
infer_quant:False
inference:tools/infer/predict_det.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
--use_tensorrt:False
---precision:fp32|int8
+--precision:fp32
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/en_server_pgnetA/train_infer_python.txt b/test_tipc/configs/en_server_pgnetA/train_infer_python.txt
index 8a1509baab..a9dd4e676b 100644
--- a/test_tipc/configs/en_server_pgnetA/train_infer_python.txt
+++ b/test_tipc/configs/en_server_pgnetA/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c configs/e2e/e2e_r50_vd_pg.yml -o
infer_quant:False
inference:tools/infer/predict_e2e.py
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1
--use_tensorrt:False
---precision:fp32|fp16|int8
+--precision:fp32
--e2e_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
diff --git a/test_tipc/configs/en_table_structure/table_mv3.yml b/test_tipc/configs/en_table_structure/table_mv3.yml
new file mode 100755
index 0000000000..adf326bd02
--- /dev/null
+++ b/test_tipc/configs/en_table_structure/table_mv3.yml
@@ -0,0 +1,117 @@
+Global:
+ use_gpu: true
+ epoch_num: 10
+ log_smooth_window: 20
+ print_batch_step: 5
+ save_model_dir: ./output/table_mv3/
+ save_epoch_step: 3
+ # evaluation is run every 400 iterations after the 0th iteration
+ eval_batch_step: [0, 400]
+ cal_metric_during_train: True
+ pretrained_model:
+ checkpoints:
+ save_inference_dir:
+ use_visualdl: False
+ infer_img: doc/table/table.jpg
+ # for data or label process
+ character_dict_path: ppocr/utils/dict/table_structure_dict.txt
+ character_type: en
+ max_text_length: 100
+ max_elem_length: 800
+ max_cell_num: 500
+ infer_mode: False
+ process_total_num: 0
+ process_cut_num: 0
+
+Optimizer:
+ name: Adam
+ beta1: 0.9
+ beta2: 0.999
+ clip_norm: 5.0
+ lr:
+ learning_rate: 0.001
+ regularizer:
+ name: 'L2'
+ factor: 0.00000
+
+Architecture:
+ model_type: table
+ algorithm: TableAttn
+ Backbone:
+ name: MobileNetV3
+ scale: 1.0
+ model_name: large
+ Head:
+ name: TableAttentionHead
+ hidden_size: 256
+ l2_decay: 0.00001
+ loc_type: 2
+ max_text_length: 100
+ max_elem_length: 800
+ max_cell_num: 500
+
+Loss:
+ name: TableAttentionLoss
+ structure_weight: 100.0
+ loc_weight: 10000.0
+
+PostProcess:
+ name: TableLabelDecode
+
+Metric:
+ name: TableMetric
+ main_indicator: acc
+
+Train:
+ dataset:
+ name: PubTabDataSet
+ data_dir: ./train_data/pubtabnet/train
+ label_file_path: ./train_data/pubtabnet/train.jsonl
+ transforms:
+ - DecodeImage: # load image
+ img_mode: BGR
+ channel_first: False
+ - ResizeTableImage:
+ max_len: 488
+ - TableLabelEncode:
+ - NormalizeImage:
+ scale: 1./255.
+ mean: [0.485, 0.456, 0.406]
+ std: [0.229, 0.224, 0.225]
+ order: 'hwc'
+ - PaddingTableImage:
+ - ToCHWImage:
+ - KeepKeys:
+ keep_keys: ['image', 'structure', 'bbox_list', 'sp_tokens', 'bbox_list_mask']
+ loader:
+ shuffle: True
+ batch_size_per_card: 32
+ drop_last: True
+ num_workers: 1
+
+Eval:
+ dataset:
+ name: PubTabDataSet
+ data_dir: ./train_data/pubtabnet/test/
+ label_file_path: ./train_data/pubtabnet/test.jsonl
+ transforms:
+ - DecodeImage: # load image
+ img_mode: BGR
+ channel_first: False
+ - ResizeTableImage:
+ max_len: 488
+ - TableLabelEncode:
+ - NormalizeImage:
+ scale: 1./255.
+ mean: [0.485, 0.456, 0.406]
+ std: [0.229, 0.224, 0.225]
+ order: 'hwc'
+ - PaddingTableImage:
+ - ToCHWImage:
+ - KeepKeys:
+ keep_keys: ['image', 'structure', 'bbox_list', 'sp_tokens', 'bbox_list_mask']
+ loader:
+ shuffle: False
+ drop_last: False
+ batch_size_per_card: 16
+ num_workers: 1
diff --git a/test_tipc/configs/en_table_structure/train_infer_python.txt b/test_tipc/configs/en_table_structure/train_infer_python.txt
new file mode 100644
index 0000000000..d9f3b30e16
--- /dev/null
+++ b/test_tipc/configs/en_table_structure/train_infer_python.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:en_table_structure
+python:python3.7
+gpu_list:0|0,1
+Global.use_gpu:True|True
+Global.auto_cast:fp32
+Global.epoch_num:lite_train_lite_infer=3|whole_train_whole_infer=50
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=128
+Global.pretrained_model:./pretrain_models/en_ppocr_mobile_v2.0_table_structure_train/best_accuracy
+train_model_name:latest
+train_infer_img_dir:./ppstructure/docs/table/table.jpg
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c test_tipc/configs/en_table_structure/table_mv3.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:null
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c test_tipc/configs/en_table_structure/table_mv3.yml -o
+quant_export:
+fpgm_export:
+distill_export:null
+export1:null
+export2:null
+##
+infer_model:./inference/en_ppocr_mobile_v2.0_table_structure_infer
+infer_export:null
+infer_quant:False
+inference:ppstructure/table/predict_table.py --det_model_dir=./inference/en_ppocr_mobile_v2.0_table_det_infer --rec_model_dir=./inference/en_ppocr_mobile_v2.0_table_rec_infer --rec_char_dict_path=./ppocr/utils/dict/table_dict.txt --table_char_dict_path=./ppocr/utils/dict/table_structure_dict.txt --image_dir=./ppstructure/docs/table/table.jpg --det_limit_side_len=736 --det_limit_type=min --output ./output/table
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--table_model_dir:
+--image_dir:./ppstructure/docs/table/table.jpg
+null:null
+--benchmark:False
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,488,488]}]
diff --git a/test_tipc/configs/en_table_structure/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/en_table_structure/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..41d236c376
--- /dev/null
+++ b/test_tipc/configs/en_table_structure/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:en_table_structure
+python:python3.7
+gpu_list:192.168.0.1,192.168.0.2;0,1
+Global.use_gpu:True
+Global.auto_cast:fp32
+Global.epoch_num:lite_train_lite_infer=3|whole_train_whole_infer=50
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=128
+Global.pretrained_model:./pretrain_models/en_ppocr_mobile_v2.0_table_structure_train/best_accuracy
+train_model_name:latest
+train_infer_img_dir:./ppstructure/docs/table/table.jpg
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c test_tipc/configs/en_table_structure/table_mv3.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:null
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c test_tipc/configs/en_table_structure/table_mv3.yml -o
+quant_export:
+fpgm_export:
+distill_export:null
+export1:null
+export2:null
+##
+infer_model:./inference/en_ppocr_mobile_v2.0_table_structure_infer
+infer_export:null
+infer_quant:False
+inference:ppstructure/table/predict_table.py --det_model_dir=./inference/en_ppocr_mobile_v2.0_table_det_infer --rec_model_dir=./inference/en_ppocr_mobile_v2.0_table_rec_infer --rec_char_dict_path=./ppocr/utils/dict/table_dict.txt --table_char_dict_path=./ppocr/utils/dict/table_structure_dict.txt --image_dir=./ppstructure/docs/table/table.jpg --det_limit_side_len=736 --det_limit_type=min --output ./output/table
+--use_gpu:False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--table_model_dir:
+--image_dir:./ppstructure/docs/table/table.jpg
+null:null
+--benchmark:False
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,488,488]}]
diff --git a/test_tipc/configs/en_table_structure/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/en_table_structure/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
new file mode 100644
index 0000000000..31ac1ed53f
--- /dev/null
+++ b/test_tipc/configs/en_table_structure/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:en_table_structure
+python:python3.7
+gpu_list:0|0,1
+Global.use_gpu:True|True
+Global.auto_cast:amp
+Global.epoch_num:lite_train_lite_infer=3|whole_train_whole_infer=50
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=128
+Global.pretrained_model:./pretrain_models/en_ppocr_mobile_v2.0_table_structure_train/best_accuracy
+train_model_name:latest
+train_infer_img_dir:./ppstructure/docs/table/table.jpg
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c test_tipc/configs/en_table_structure/table_mv3.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:null
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c test_tipc/configs/en_table_structure/table_mv3.yml -o
+quant_export:
+fpgm_export:
+distill_export:null
+export1:null
+export2:null
+##
+infer_model:./inference/en_ppocr_mobile_v2.0_table_structure_infer
+infer_export:null
+infer_quant:False
+inference:ppstructure/table/predict_table.py --det_model_dir=./inference/en_ppocr_mobile_v2.0_table_det_infer --rec_model_dir=./inference/en_ppocr_mobile_v2.0_table_rec_infer --rec_char_dict_path=./ppocr/utils/dict/table_dict.txt --table_char_dict_path=./ppocr/utils/dict/table_structure_dict.txt --image_dir=./ppstructure/docs/table/table.jpg --det_limit_side_len=736 --det_limit_type=min --output ./output/table
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--table_model_dir:
+--image_dir:./ppstructure/docs/table/table.jpg
+null:null
+--benchmark:False
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,488,488]}]
diff --git a/test_tipc/configs/en_table_structure/train_pact_infer_python.txt b/test_tipc/configs/en_table_structure/train_pact_infer_python.txt
new file mode 100644
index 0000000000..f62e8b68bc
--- /dev/null
+++ b/test_tipc/configs/en_table_structure/train_pact_infer_python.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:en_table_structure_PACT
+python:python3.7
+gpu_list:0|0,1
+Global.use_gpu:True|True
+Global.auto_cast:fp32
+Global.epoch_num:lite_train_lite_infer=1|whole_train_whole_infer=50
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=128
+Global.pretrained_model:./pretrain_models/en_ppocr_mobile_v2.0_table_structure_train/best_accuracy
+train_model_name:latest
+train_infer_img_dir:./ppstructure/docs/table/table.jpg
+null:null
+##
+trainer:pact_train
+norm_train:null
+pact_train:deploy/slim/quantization/quant.py -c test_tipc/configs/en_table_structure/table_mv3.yml -o
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:null
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:null
+quant_export:deploy/slim/quantization/export_model.py -c test_tipc/configs/en_table_structure/table_mv3.yml -o
+fpgm_export:
+distill_export:null
+export1:null
+export2:null
+##
+infer_model:./inference/en_ppocr_mobile_v2.0_table_structure_infer
+infer_export:null
+infer_quant:True
+inference:ppstructure/table/predict_table.py --det_model_dir=./inference/en_ppocr_mobile_v2.0_table_det_infer --rec_model_dir=./inference/en_ppocr_mobile_v2.0_table_rec_infer --rec_char_dict_path=./ppocr/utils/dict/table_dict.txt --table_char_dict_path=./ppocr/utils/dict/table_structure_dict.txt --image_dir=./ppstructure/docs/table/table.jpg --det_limit_side_len=736 --det_limit_type=min --output ./output/table
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:fp32
+--table_model_dir:
+--image_dir:./ppstructure/docs/table/table.jpg
+null:null
+--benchmark:False
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,488,488]}]
diff --git a/test_tipc/configs/en_table_structure/train_ptq_infer_python.txt b/test_tipc/configs/en_table_structure/train_ptq_infer_python.txt
new file mode 100644
index 0000000000..e8f7bbaa50
--- /dev/null
+++ b/test_tipc/configs/en_table_structure/train_ptq_infer_python.txt
@@ -0,0 +1,21 @@
+===========================train_params===========================
+model_name:en_table_structure_KL
+python:python3.7
+Global.pretrained_model:
+Global.save_inference_dir:null
+infer_model:./inference/en_ppocr_mobile_v2.0_table_structure_infer/
+infer_export:deploy/slim/quantization/quant_kl.py -c test_tipc/configs/en_table_structure/table_mv3.yml -o
+infer_quant:True
+inference:ppstructure/table/predict_table.py --det_model_dir=./inference/en_ppocr_mobile_v2.0_table_det_infer --rec_model_dir=./inference/en_ppocr_mobile_v2.0_table_rec_infer --rec_char_dict_path=./ppocr/utils/dict/table_dict.txt --table_char_dict_path=./ppocr/utils/dict/table_structure_dict.txt --image_dir=./ppstructure/docs/table/table.jpg --det_limit_side_len=736 --det_limit_type=min --output ./output/table
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1
+--use_tensorrt:False
+--precision:int8
+--table_model_dir:
+--image_dir:./ppstructure/docs/table/table.jpg
+null:null
+--benchmark:False
+null:null
+null:null
diff --git a/test_tipc/configs/rec_mtb_nrtr/rec_mtb_nrtr.yml b/test_tipc/configs/rec_mtb_nrtr/rec_mtb_nrtr.yml
index 15119bb2a9..8118d58724 100644
--- a/test_tipc/configs/rec_mtb_nrtr/rec_mtb_nrtr.yml
+++ b/test_tipc/configs/rec_mtb_nrtr/rec_mtb_nrtr.yml
@@ -49,7 +49,7 @@ Architecture:
Loss:
- name: NRTRLoss
+ name: CELoss
smoothing: True
PostProcess:
@@ -69,7 +69,7 @@ Train:
img_mode: BGR
channel_first: False
- NRTRLabelEncode: # Class handling label
- - NRTRRecResizeImg:
+ - GrayRecResizeImg:
image_shape: [100, 32]
resize_type: PIL # PIL or OpenCV
- KeepKeys:
@@ -90,7 +90,7 @@ Eval:
img_mode: BGR
channel_first: False
- NRTRLabelEncode: # Class handling label
- - NRTRRecResizeImg:
+ - GrayRecResizeImg:
image_shape: [100, 32]
resize_type: PIL # PIL or OpenCV
- KeepKeys:
@@ -99,5 +99,5 @@ Eval:
shuffle: False
drop_last: False
batch_size_per_card: 256
- num_workers: 1
+ num_workers: 4
use_shared_memory: False
diff --git a/test_tipc/configs/rec_mtb_nrtr/train_infer_python.txt b/test_tipc/configs/rec_mtb_nrtr/train_infer_python.txt
index de6de5a0ca..fed8ba2675 100644
--- a/test_tipc/configs/rec_mtb_nrtr/train_infer_python.txt
+++ b/test_tipc/configs/rec_mtb_nrtr/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/rec_mtb_nrtr/rec_mtb_nrt
infer_quant:False
inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/EN_symbol_dict.txt --rec_image_shape="1,32,100" --rec_algorithm="NRTR"
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/train_infer_python.txt b/test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/train_infer_python.txt
index e67dd15090..39bf922790 100644
--- a/test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/train_infer_python.txt
+++ b/test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/rec_mv3_none_bilstm_ctc_
infer_quant:False
inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,100"
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/rec_mv3_none_none_ctc_v2.0/train_infer_python.txt b/test_tipc/configs/rec_mv3_none_none_ctc_v2.0/train_infer_python.txt
index aa3e88d284..593de3ff20 100644
--- a/test_tipc/configs/rec_mv3_none_none_ctc_v2.0/train_infer_python.txt
+++ b/test_tipc/configs/rec_mv3_none_none_ctc_v2.0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/rec_mv3_none_none_ctc_v2
infer_quant:False
inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,100"
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/rec_mv3_tps_bilstm_att_v2.0/train_infer_python.txt b/test_tipc/configs/rec_mv3_tps_bilstm_att_v2.0/train_infer_python.txt
index c22767c60f..1b2d9abb0f 100644
--- a/test_tipc/configs/rec_mv3_tps_bilstm_att_v2.0/train_infer_python.txt
+++ b/test_tipc/configs/rec_mv3_tps_bilstm_att_v2.0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/rec_mv3_tps_bilstm_att_v
infer_quant:False
inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,100" --rec_algorithm="RARE" --min_subgraph_size=5
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/train_infer_python.txt b/test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/train_infer_python.txt
index 7a3096eb1e..1367c7abd4 100644
--- a/test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/train_infer_python.txt
+++ b/test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/rec_mv3_tps_bilstm_ctc_v
infer_quant:False
inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,100" --rec_algorithm="StarNet"
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/rec_r31_sar/train_infer_python.txt b/test_tipc/configs/rec_r31_sar/train_infer_python.txt
index 1a32a3d507..03ec54abb6 100644
--- a/test_tipc/configs/rec_r31_sar/train_infer_python.txt
+++ b/test_tipc/configs/rec_r31_sar/train_infer_python.txt
@@ -38,12 +38,12 @@ train_model:./inference/rec_r31_sar_train/best_accuracy
infer_export:tools/export_model.py -c test_tipc/configs/rec_r31_sar/rec_r31_sar.yml -o
infer_quant:False
inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/dict90.txt --rec_image_shape="3,48,48,160" --rec_algorithm="SAR"
---use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--use_gpu:True
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/train_infer_python.txt b/test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/train_infer_python.txt
index 02cea56fbe..46aa3d7190 100644
--- a/test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/train_infer_python.txt
+++ b/test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/rec_r34_vd_none_bilstm_c
infer_quant:False
inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,100"
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/train_infer_python.txt b/test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/train_infer_python.txt
index 5e7c1d3431..3e066d7b72 100644
--- a/test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/train_infer_python.txt
+++ b/test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/rec_r34_vd_none_none_ctc
infer_quant:False
inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,100"
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/rec_r34_vd_tps_bilstm_att_v2.0/train_infer_python.txt b/test_tipc/configs/rec_r34_vd_tps_bilstm_att_v2.0/train_infer_python.txt
index 55e937881b..1e4f46633e 100644
--- a/test_tipc/configs/rec_r34_vd_tps_bilstm_att_v2.0/train_infer_python.txt
+++ b/test_tipc/configs/rec_r34_vd_tps_bilstm_att_v2.0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/rec_r34_vd_tps_bilstm_at
infer_quant:False
inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,100" --rec_algorithm="RARE" --min_subgraph_size=5
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/train_infer_python.txt b/test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/train_infer_python.txt
index 5b5ba0fd01..9e795b6645 100644
--- a/test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/train_infer_python.txt
+++ b/test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/rec_r34_vd_tps_bilstm_ct
infer_quant:False
inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,100" --rec_algorithm="StarNet"
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/rec_r45_abinet/rec_r45_abinet.yml b/test_tipc/configs/rec_r45_abinet/rec_r45_abinet.yml
new file mode 100644
index 0000000000..5b5890e772
--- /dev/null
+++ b/test_tipc/configs/rec_r45_abinet/rec_r45_abinet.yml
@@ -0,0 +1,106 @@
+Global:
+ use_gpu: True
+ epoch_num: 10
+ log_smooth_window: 20
+ print_batch_step: 10
+ save_model_dir: ./output/rec/r45_abinet/
+ save_epoch_step: 1
+ # evaluation is run every 2000 iterations
+ eval_batch_step: [0, 2000]
+ cal_metric_during_train: True
+ pretrained_model:
+ checkpoints:
+ save_inference_dir:
+ use_visualdl: False
+ infer_img: doc/imgs_words_en/word_10.png
+ # for data or label process
+ character_dict_path:
+ character_type: en
+ max_text_length: 25
+ infer_mode: False
+ use_space_char: False
+ save_res_path: ./output/rec/predicts_abinet.txt
+
+Optimizer:
+ name: Adam
+ beta1: 0.9
+ beta2: 0.99
+ clip_norm: 20.0
+ lr:
+ name: Piecewise
+ decay_epochs: [6]
+ values: [0.0001, 0.00001]
+ regularizer:
+ name: 'L2'
+ factor: 0.
+
+Architecture:
+ model_type: rec
+ algorithm: ABINet
+ in_channels: 3
+ Transform:
+ Backbone:
+ name: ResNet45
+
+ Head:
+ name: ABINetHead
+ use_lang: True
+ iter_size: 3
+
+
+Loss:
+ name: CELoss
+ ignore_index: &ignore_index 100 # Must be greater than the number of character classes
+
+PostProcess:
+ name: ABINetLabelDecode
+
+Metric:
+ name: RecMetric
+ main_indicator: acc
+
+Train:
+ dataset:
+ name: SimpleDataSet
+ data_dir: ./train_data/ic15_data/
+ label_file_list: ["./train_data/ic15_data/rec_gt_train.txt"]
+ transforms:
+ - DecodeImage: # load image
+ img_mode: RGB
+ channel_first: False
+ - ABINetRecAug:
+ - ABINetLabelEncode: # Class handling label
+ ignore_index: *ignore_index
+ - ABINetRecResizeImg:
+ image_shape: [3, 32, 128]
+ padding: False
+ - KeepKeys:
+ keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order
+ loader:
+ shuffle: True
+ batch_size_per_card: 96
+ drop_last: True
+ num_workers: 4
+
+Eval:
+ dataset:
+ name: SimpleDataSet
+ data_dir: ./train_data/ic15_data
+ label_file_list: ["./train_data/ic15_data/rec_gt_test.txt"]
+ transforms:
+ - DecodeImage: # load image
+ img_mode: RGB
+ channel_first: False
+ - ABINetLabelEncode: # Class handling label
+ ignore_index: *ignore_index
+ - ABINetRecResizeImg:
+ image_shape: [3, 32, 128]
+ 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
+ use_shared_memory: False
diff --git a/test_tipc/configs/rec_r45_abinet/train_infer_python.txt b/test_tipc/configs/rec_r45_abinet/train_infer_python.txt
new file mode 100644
index 0000000000..ecab1bcbbd
--- /dev/null
+++ b/test_tipc/configs/rec_r45_abinet/train_infer_python.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:rec_abinet
+python:python3.7
+gpu_list:0|0,1
+Global.use_gpu:True|True
+Global.auto_cast:null
+Global.epoch_num:lite_train_lite_infer=2|whole_train_whole_infer=300
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=64
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./inference/rec_inference
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c test_tipc/configs/rec_r45_abinet/rec_r45_abinet.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:tools/eval.py -c test_tipc/configs/rec_r45_abinet/rec_r45_abinet.yml -o
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c test_tipc/configs/rec_r45_abinet/rec_r45_abinet.yml -o
+quant_export:null
+fpgm_export:null
+distill_export:null
+export1:null
+export2:null
+##
+train_model:./inference/rec_r45_abinet_train/best_accuracy
+infer_export:tools/export_model.py -c test_tipc/configs/rec_r45_abinet/rec_r45_abinet.yml -o
+infer_quant:False
+inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,128" --rec_algorithm="ABINet"
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1|6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference
+--save_log_path:./test/output/
+--benchmark:True
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,32,128]}]
diff --git a/test_tipc/configs/rec_r50_fpn_vd_none_srn/train_infer_python.txt b/test_tipc/configs/rec_r50_fpn_vd_none_srn/train_infer_python.txt
index 4877512b68..b5a5286010 100644
--- a/test_tipc/configs/rec_r50_fpn_vd_none_srn/train_infer_python.txt
+++ b/test_tipc/configs/rec_r50_fpn_vd_none_srn/train_infer_python.txt
@@ -39,11 +39,11 @@ infer_export:tools/export_model.py -c test_tipc/configs/rec_r50_fpn_vd_none_srn/
infer_quant:False
inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="1,64,256" --rec_algorithm="SRN" --use_space_char=False --min_subgraph_size=3
--use_gpu:True|False
---enable_mkldnn:True|False
---cpu_threads:1|6
+--enable_mkldnn:False
+--cpu_threads:6
--rec_batch_num:1|6
---use_tensorrt:True|False
---precision:fp32|int8
+--use_tensorrt:False
+--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference
--save_log_path:./test/output/
diff --git a/test_tipc/configs/rec_svtrnet/rec_svtrnet.yml b/test_tipc/configs/rec_svtrnet/rec_svtrnet.yml
new file mode 100644
index 0000000000..140b17e0e7
--- /dev/null
+++ b/test_tipc/configs/rec_svtrnet/rec_svtrnet.yml
@@ -0,0 +1,117 @@
+Global:
+ use_gpu: True
+ epoch_num: 20
+ log_smooth_window: 20
+ print_batch_step: 10
+ save_model_dir: ./output/rec/svtr/
+ save_epoch_step: 1
+ # evaluation is run every 2000 iterations after the 0th iteration
+ eval_batch_step: [0, 2000]
+ cal_metric_during_train: True
+ pretrained_model:
+ checkpoints:
+ save_inference_dir:
+ use_visualdl: False
+ infer_img: doc/imgs_words_en/word_10.png
+ # for data or label process
+ character_dict_path:
+ character_type: en
+ max_text_length: 25
+ infer_mode: False
+ use_space_char: False
+ save_res_path: ./output/rec/predicts_svtr_tiny.txt
+
+
+Optimizer:
+ name: AdamW
+ beta1: 0.9
+ beta2: 0.99
+ epsilon: 8.e-8
+ weight_decay: 0.05
+ no_weight_decay_name: norm pos_embed
+ one_dim_param_no_weight_decay: true
+ lr:
+ name: Cosine
+ learning_rate: 0.0005
+ warmup_epoch: 2
+
+Architecture:
+ model_type: rec
+ algorithm: SVTR
+ Transform:
+ name: STN_ON
+ tps_inputsize: [32, 64]
+ tps_outputsize: [32, 100]
+ num_control_points: 20
+ tps_margins: [0.05,0.05]
+ stn_activation: none
+ Backbone:
+ name: SVTRNet
+ img_size: [32, 100]
+ out_char_num: 25
+ out_channels: 192
+ patch_merging: 'Conv'
+ embed_dim: [64, 128, 256]
+ depth: [3, 6, 3]
+ num_heads: [2, 4, 8]
+ mixer: ['Local','Local','Local','Local','Local','Local','Global','Global','Global','Global','Global','Global']
+ local_mixer: [[7, 11], [7, 11], [7, 11]]
+ last_stage: True
+ prenorm: false
+ Neck:
+ name: SequenceEncoder
+ encoder_type: reshape
+ Head:
+ name: CTCHead
+
+Loss:
+ name: CTCLoss
+
+PostProcess:
+ name: CTCLabelDecode
+
+Metric:
+ name: RecMetric
+ main_indicator: acc
+
+Train:
+ dataset:
+ name: SimpleDataSet
+ data_dir: ./train_data/ic15_data/
+ label_file_list: ["./train_data/ic15_data/rec_gt_train.txt"]
+ transforms:
+ - DecodeImage: # load image
+ img_mode: BGR
+ channel_first: False
+ - CTCLabelEncode: # Class handling label
+ - SVTRRecResizeImg:
+ image_shape: [3, 64, 256]
+ padding: False
+ - KeepKeys:
+ keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order
+ loader:
+ shuffle: True
+ batch_size_per_card: 512
+ drop_last: True
+ num_workers: 4
+
+Eval:
+ dataset:
+ name: SimpleDataSet
+ data_dir: ./train_data/ic15_data
+ label_file_list: ["./train_data/ic15_data/rec_gt_test.txt"]
+ transforms:
+ - DecodeImage: # load image
+ img_mode: BGR
+ channel_first: False
+ - CTCLabelEncode: # Class handling label
+ - SVTRRecResizeImg:
+ image_shape: [3, 64, 256]
+ 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: 2
diff --git a/test_tipc/configs/rec_svtrnet/train_infer_python.txt b/test_tipc/configs/rec_svtrnet/train_infer_python.txt
new file mode 100644
index 0000000000..a7e4a24063
--- /dev/null
+++ b/test_tipc/configs/rec_svtrnet/train_infer_python.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:rec_svtrnet
+python:python3.7
+gpu_list:0|0,1
+Global.use_gpu:True|True
+Global.auto_cast:null
+Global.epoch_num:lite_train_lite_infer=2|whole_train_whole_infer=300
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=64
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./inference/rec_inference
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c test_tipc/configs/rec_svtrnet/rec_svtrnet.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:tools/eval.py -c test_tipc/configs/rec_svtrnet/rec_svtrnet.yml -o
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c test_tipc/configs/rec_svtrnet/rec_svtrnet.yml -o
+quant_export:null
+fpgm_export:null
+distill_export:null
+export1:null
+export2:null
+##
+train_model:./inference/rec_svtrnet_train/best_accuracy
+infer_export:tools/export_model.py -c test_tipc/configs/rec_svtrnet/rec_svtrnet.yml -o
+infer_quant:False
+inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,64,256" --rec_algorithm="SVTR"
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1|6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference
+--save_log_path:./test/output/
+--benchmark:True
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,64,256]}]
diff --git a/test_tipc/configs/rec_vitstr_none_ce/rec_vitstr_none_ce.yml b/test_tipc/configs/rec_vitstr_none_ce/rec_vitstr_none_ce.yml
new file mode 100644
index 0000000000..a0aed48875
--- /dev/null
+++ b/test_tipc/configs/rec_vitstr_none_ce/rec_vitstr_none_ce.yml
@@ -0,0 +1,104 @@
+Global:
+ use_gpu: True
+ epoch_num: 20
+ log_smooth_window: 20
+ print_batch_step: 10
+ save_model_dir: ./output/rec/vitstr_none_ce/
+ save_epoch_step: 1
+ # evaluation is run every 2000 iterations after the 0th iteration#
+ eval_batch_step: [0, 2000]
+ cal_metric_during_train: True
+ pretrained_model:
+ checkpoints:
+ save_inference_dir:
+ use_visualdl: False
+ infer_img: doc/imgs_words_en/word_10.png
+ # for data or label process
+ character_dict_path: ppocr/utils/EN_symbol_dict.txt
+ max_text_length: 25
+ infer_mode: False
+ use_space_char: False
+ save_res_path: ./output/rec/predicts_vitstr.txt
+
+
+Optimizer:
+ name: Adadelta
+ epsilon: 1.e-8
+ rho: 0.95
+ clip_norm: 5.0
+ lr:
+ learning_rate: 1.0
+
+Architecture:
+ model_type: rec
+ algorithm: ViTSTR
+ in_channels: 1
+ Transform:
+ Backbone:
+ name: ViTSTR
+ Neck:
+ name: SequenceEncoder
+ encoder_type: reshape
+ Head:
+ name: CTCHead
+
+Loss:
+ name: CELoss
+ smoothing: False
+ with_all: True
+ ignore_index: &ignore_index 0 # Must be zero or greater than the number of character classes
+
+PostProcess:
+ name: ViTSTRLabelDecode
+
+Metric:
+ name: RecMetric
+ main_indicator: acc
+
+Train:
+ dataset:
+ name: SimpleDataSet
+ data_dir: ./train_data/ic15_data/
+ label_file_list: ["./train_data/ic15_data/rec_gt_train.txt"]
+ transforms:
+ - DecodeImage: # load image
+ img_mode: BGR
+ channel_first: False
+ - ViTSTRLabelEncode: # Class handling label
+ ignore_index: *ignore_index
+ - GrayRecResizeImg:
+ image_shape: [224, 224] # W H
+ resize_type: PIL # PIL or OpenCV
+ inter_type: 'Image.BICUBIC'
+ scale: false
+ - KeepKeys:
+ keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order
+ loader:
+ shuffle: True
+ batch_size_per_card: 48
+ drop_last: True
+ num_workers: 8
+
+Eval:
+ dataset:
+ name: SimpleDataSet
+ data_dir: ./train_data/ic15_data
+ label_file_list: ["./train_data/ic15_data/rec_gt_test.txt"]
+ transforms:
+ - DecodeImage: # load image
+ img_mode: BGR
+ channel_first: False
+ - ViTSTRLabelEncode: # Class handling label
+ ignore_index: *ignore_index
+ - GrayRecResizeImg:
+ image_shape: [224, 224] # W H
+ resize_type: PIL # PIL or OpenCV
+ inter_type: 'Image.BICUBIC'
+ scale: 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: 2
diff --git a/test_tipc/configs/rec_vitstr_none_ce/train_infer_python.txt b/test_tipc/configs/rec_vitstr_none_ce/train_infer_python.txt
new file mode 100644
index 0000000000..04c5742ea2
--- /dev/null
+++ b/test_tipc/configs/rec_vitstr_none_ce/train_infer_python.txt
@@ -0,0 +1,53 @@
+===========================train_params===========================
+model_name:rec_vitstr
+python:python3.7
+gpu_list:0|0,1
+Global.use_gpu:True|True
+Global.auto_cast:null
+Global.epoch_num:lite_train_lite_infer=2|whole_train_whole_infer=300
+Global.save_model_dir:./output/
+Train.loader.batch_size_per_card:lite_train_lite_infer=16|whole_train_whole_infer=64
+Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./inference/rec_inference
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c test_tipc/configs/rec_vitstr_none_ce/rec_vitstr_none_ce.yml -o
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:tools/eval.py -c test_tipc/configs/rec_vitstr_none_ce/rec_vitstr_none_ce.yml -o
+null:null
+##
+===========================infer_params===========================
+Global.save_inference_dir:./output/
+Global.checkpoints:
+norm_export:tools/export_model.py -c test_tipc/configs/rec_vitstr_none_ce/rec_vitstr_none_ce.yml -o
+quant_export:null
+fpgm_export:null
+distill_export:null
+export1:null
+export2:null
+##
+train_model:./inference/rec_vitstr_none_ce_train/best_accuracy
+infer_export:tools/export_model.py -c test_tipc/configs/rec_vitstr_none_ce/rec_vitstr_none_ce.yml -o
+infer_quant:False
+inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/EN_symbol_dict.txt --rec_image_shape="1,224,224" --rec_algorithm="ViTSTR"
+--use_gpu:True|False
+--enable_mkldnn:False
+--cpu_threads:6
+--rec_batch_num:1|6
+--use_tensorrt:False
+--precision:fp32
+--rec_model_dir:
+--image_dir:./inference/rec_inference
+--save_log_path:./test/output/
+--benchmark:True
+null:null
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[1,224,224]}]
diff --git a/test_tipc/docs/jeston_test_train_inference_python.md b/test_tipc/docs/jeston_test_train_inference_python.md
index 9e9d15fb67..b25175ed00 100644
--- a/test_tipc/docs/jeston_test_train_inference_python.md
+++ b/test_tipc/docs/jeston_test_train_inference_python.md
@@ -115,4 +115,4 @@ ValueError: The results of python_infer_gpu_usetrt_True_precision_fp32_batchsize
## 3. 更多教程
本文档为功能测试用，更丰富的训练预测使用教程请参考：
[模型训练](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/training.md)
-[基于Python预测引擎推理](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/inference.md)
+[基于Python预测引擎推理](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/inference_ppocr.md)
diff --git a/test_tipc/docs/mac_test_train_inference_python.md b/test_tipc/docs/mac_test_train_inference_python.md
index ea6e0218b1..c37291a8fc 100644
--- a/test_tipc/docs/mac_test_train_inference_python.md
+++ b/test_tipc/docs/mac_test_train_inference_python.md
@@ -152,4 +152,4 @@ ValueError: The results of python_infer_cpu_usemkldnn_False_threads_1_batchsize_
## 3. 更多教程
本文档为功能测试用，更丰富的训练预测使用教程请参考：
[模型训练](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/training.md)
-[基于Python预测引擎推理](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/inference.md)
+[基于Python预测引擎推理](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/inference_ppocr.md)
diff --git a/test_tipc/docs/test_serving.md b/test_tipc/docs/test_serving.md
index 8600eff3b9..71f01c0d5f 100644
--- a/test_tipc/docs/test_serving.md
+++ b/test_tipc/docs/test_serving.md
@@ -1,6 +1,6 @@
# PaddleServing预测功能测试
-PaddleServing预测功能测试的主程序为`test_serving.sh`，可以测试基于PaddleServing的部署功能。
+PaddleServing预测功能测试的主程序为`test_serving_infer_python.sh`和`test_serving_infer_cpp.sh`，可以测试基于PaddleServing的部署功能。
## 1. 测试结论汇总
@@ -17,13 +17,23 @@ PaddleServing预测功能测试的主程序为`test_serving.sh`，可以测试
运行环境配置请参考[文档](./install.md)的内容配置TIPC的运行环境。
### 2.1 功能测试
-先运行`prepare.sh`准备数据和模型，然后运行`test_serving.sh`进行测试，最终在```test_tipc/output```目录下生成`serving_infer_*.log`后缀的日志文件。
+**python serving**
+先运行`prepare.sh`准备数据和模型，然后运行`test_serving_infer_python.sh`进行测试，最终在```test_tipc/output```目录下生成`serving_infer_python*.log`后缀的日志文件。
```shell
bash test_tipc/prepare.sh ./test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt "serving_infer"
# 用法:
-bash test_tipc/test_serving.sh ./test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
+bash test_tipc/test_serving_infer_python.sh ./test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt "serving_infer"
+```
+**cpp serving**
+先运行`prepare.sh`准备数据和模型，然后运行`test_serving_infer_cpp.sh`进行测试，最终在```test_tipc/output```目录下生成`serving_infer_cpp*.log`后缀的日志文件。
+
+```shell
+bash test_tipc/prepare.sh ./test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt "serving_infer"
+
+# 用法:
+bash test_tipc/test_serving_infer_cpp.sh ./test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt "serving_infer"
```
#### 运行结果
diff --git a/test_tipc/docs/test_train_fleet_inference_python.md b/test_tipc/docs/test_train_fleet_inference_python.md
new file mode 100644
index 0000000000..9fddb5d163
--- /dev/null
+++ b/test_tipc/docs/test_train_fleet_inference_python.md
@@ -0,0 +1,107 @@
+# Linux GPU/CPU 多机多卡训练推理测试
+
+Linux GPU/CPU 多机多卡训练推理测试的主程序为`test_train_inference_python.sh`，可以测试基于Python的模型训练、评估、推理等基本功能。
+
+## 1. 测试结论汇总
+
+- 训练相关：
+
+| 算法名称 | 模型名称 | 多机多卡 |
+| :----: | :----: | :----: |
+| PP-OCRv3 | ch_PP-OCRv3_rec | 分布式训练 |
+
+
+- 推理相关：
+
+| 算法名称 | 模型名称 | device_CPU | device_GPU | batchsize |
+| :----: | :----: | :----: | :----: | :----: |
+| PP-OCRv3 | ch_PP-OCRv3_rec | 支持 | - | 1/6 |
+
+
+## 2. 测试流程
+
+运行环境配置请参考[文档](./install.md)的内容配置TIPC的运行环境。
+
+### 2.1 功能测试
+
+#### 2.1.1 修改配置文件
+
+首先，修改配置文件中的`ip`设置: 假设两台机器的`ip`地址分别为`192.168.0.1`和`192.168.0.2`，则对应的配置文件`gpu_list`字段需要修改为`gpu_list:192.168.0.1,192.168.0.2;0,1`； `ip`地址查看命令为`ifconfig`。
+
+
+#### 2.1.2 准备数据
+
+运行`prepare.sh`准备数据和模型，以配置文件`test_tipc/configs/ch_PP-OCRv3_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt`为例，数据准备命令如下所示。
+
+```shell
+bash test_tipc/prepare.sh test_tipc/configs/ch_PP-OCRv3_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt lite_train_lite_infer
+```
+
+**注意：** 由于是多机训练，这里需要在所有的节点上均运行启动上述命令，准备数据。
+
+#### 2.1.3 修改起始端口并开始测试
+
+在多机的节点上使用下面的命令设置分布式的起始端口（否则后面运行的时候会由于无法找到运行端口而hang住），一般建议设置在`10000~20000`之间。
+
+```shell
+export FLAGS_START_PORT=17000
+```
+
+以配置文件`test_tipc/configs/ch_PP-OCRv3_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt`为例，测试方法如下所示。
+
+```shell
+bash test_tipc/test_train_inference_python.sh test_tipc/configs/ch_PP-OCRv3_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt lite_train_lite_infer
+```
+
+**注意：** 由于是多机训练，这里需要在所有的节点上均运行启动上述命令进行测试。
+
+
+#### 2.1.4 输出结果
+
+输出结果如下，表示命令运行成功。
+
+```bash
+ Run successfully with command - ch_PP-OCRv3_rec - python3.7 -m paddle.distributed.launch --ips=192.168.0.1,192.168.0.2 --gpus=0,1 tools/train.py -c test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml -o Global.use_gpu=True Global.save_model_dir=./test_tipc/output/ch_PP-OCRv3_rec/lite_train_lite_infer/norm_train_gpus_0,1_autocast_fp32_nodes_2 Global.epoch_num=3 Global.auto_cast=fp32 Train.loader.batch_size_per_card=16 !
+ ......
+ Run successfully with command - ch_PP-OCRv3_rec - python3.7 tools/infer/predict_rec.py --rec_image_shape="3,48,320" --use_gpu=False --enable_mkldnn=False --cpu_threads=6 --rec_model_dir=./test_tipc/output/ch_PP-OCRv3_rec/lite_train_lite_infer/norm_train_gpus_0,1_autocast_fp32_nodes_2/Student --rec_batch_num=1 --image_dir=./inference/rec_inference --benchmark=True --precision=fp32 > ./test_tipc/output/ch_PP-OCRv3_rec/lite_train_lite_infer/python_infer_cpu_usemkldnn_False_threads_6_precision_fp32_batchsize_1.log 2>&1 !
+```
+
+在开启benchmark参数时，可以得到测试的详细数据，包含运行环境信息（系统版本、CUDA版本、CUDNN版本、驱动版本），Paddle版本信息，参数设置信息（运行设备、线程数、是否开启内存优化等），模型信息（模型名称、精度），数据信息（batchsize、是否为动态shape等），性能信息（CPU,GPU的占用、运行耗时、预处理耗时、推理耗时、后处理耗时），内容如下所示：
+
+```
+[2022/06/02 22:53:35] ppocr INFO:
+
+[2022/06/02 22:53:35] ppocr INFO: ---------------------- Env info ----------------------
+[2022/06/02 22:53:35] ppocr INFO: OS_version: Ubuntu 16.04
+[2022/06/02 22:53:35] ppocr INFO: CUDA_version: 10.1.243
+[2022/06/02 22:53:35] ppocr INFO: CUDNN_version: 7.6.5
+[2022/06/02 22:53:35] ppocr INFO: drivier_version: 460.32.03
+[2022/06/02 22:53:35] ppocr INFO: ---------------------- Paddle info ----------------------
+[2022/06/02 22:53:35] ppocr INFO: paddle_version: 2.3.0-rc0
+[2022/06/02 22:53:35] ppocr INFO: paddle_commit: 5d4980c052583fec022812d9c29460aff7cdc18b
+[2022/06/02 22:53:35] ppocr INFO: log_api_version: 1.0
+[2022/06/02 22:53:35] ppocr INFO: ----------------------- Conf info -----------------------
+[2022/06/02 22:53:35] ppocr INFO: runtime_device: cpu
+[2022/06/02 22:53:35] ppocr INFO: ir_optim: True
+[2022/06/02 22:53:35] ppocr INFO: enable_memory_optim: True
+[2022/06/02 22:53:35] ppocr INFO: enable_tensorrt: False
+[2022/06/02 22:53:35] ppocr INFO: enable_mkldnn: False
+[2022/06/02 22:53:35] ppocr INFO: cpu_math_library_num_threads: 6
+[2022/06/02 22:53:35] ppocr INFO: ----------------------- Model info ----------------------
+[2022/06/02 22:53:35] ppocr INFO: model_name: rec
+[2022/06/02 22:53:35] ppocr INFO: precision: fp32
+[2022/06/02 22:53:35] ppocr INFO: ----------------------- Data info -----------------------
+[2022/06/02 22:53:35] ppocr INFO: batch_size: 1
+[2022/06/02 22:53:35] ppocr INFO: input_shape: dynamic
+[2022/06/02 22:53:35] ppocr INFO: data_num: 6
+[2022/06/02 22:53:35] ppocr INFO: ----------------------- Perf info -----------------------
+[2022/06/02 22:53:35] ppocr INFO: cpu_rss(MB): 288.957, gpu_rss(MB): None, gpu_util: None%
+[2022/06/02 22:53:35] ppocr INFO: total time spent(s): 0.4824
+[2022/06/02 22:53:35] ppocr INFO: preprocess_time(ms): 0.1136, inference_time(ms): 79.5877, postprocess_time(ms): 0.6945
+```
+
+该信息可以在运行log中查看，以上面的`ch_PP-OCRv3_rec`为例，log位置在`./test_tipc/output/ch_PP-OCRv3_rec/lite_train_lite_infer/results_python.log`。
+
+如果运行失败，也会在终端中输出运行失败的日志信息以及对应的运行命令。可以基于该命令，分析运行失败的原因。
+
+**注意：** 由于分布式训练时，仅在`trainer_id=0`所在的节点中保存模型，因此其他的节点中在运行模型导出与推理时会报错，为正常现象。
diff --git a/test_tipc/docs/test_train_inference_python.md b/test_tipc/docs/test_train_inference_python.md
index fa969cbe1b..99de940079 100644
--- a/test_tipc/docs/test_train_inference_python.md
+++ b/test_tipc/docs/test_train_inference_python.md
@@ -153,4 +153,4 @@ python3.7 test_tipc/compare_results.py --gt_file=./test_tipc/results/python_*.tx
## 3. 更多教程
本文档为功能测试用，更丰富的训练预测使用教程请参考：
[模型训练](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/training.md)
-[基于Python预测引擎推理](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/inference.md)
+[基于Python预测引擎推理](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/inference_ppocr.md)
diff --git a/test_tipc/docs/win_test_train_inference_python.md b/test_tipc/docs/win_test_train_inference_python.md
index 95585af038..6e3ce93bb3 100644
--- a/test_tipc/docs/win_test_train_inference_python.md
+++ b/test_tipc/docs/win_test_train_inference_python.md
@@ -156,4 +156,4 @@ ValueError: The results of python_infer_cpu_usemkldnn_False_threads_1_batchsize_
## 3. 更多教程
本文档为功能测试用，更丰富的训练预测使用教程请参考：
[模型训练](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/training.md)
-[基于Python预测引擎推理](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/inference.md)
+[基于Python预测引擎推理](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/inference_ppocr.md)
diff --git a/test_tipc/prepare.sh b/test_tipc/prepare.sh
index 6a8983009e..2c9bd2901b 100644
--- a/test_tipc/prepare.sh
+++ b/test_tipc/prepare.sh
@@ -44,30 +44,49 @@ if [ ${MODE} = "lite_train_lite_infer" ];then
# pretrain lite train data
wget -nc -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/MobileNetV3_large_x0_5_pretrained.pdparams --no-check-certificate
wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_mv3_db_v2.0_train.tar --no-check-certificate
- if [[ ${model_name} =~ "PPOCRv2_det" ]];then
+ if [[ ${model_name} =~ "ch_PP-OCRv2_det" ]];then
wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_distill_train.tar --no-check-certificate
cd ./pretrain_models/ && tar xf ch_PP-OCRv2_det_distill_train.tar && cd ../
fi
+ if [[ ${model_name} =~ "ch_PP-OCRv3_det" ]];then
+ wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_distill_train.tar --no-check-certificate
+ cd ./pretrain_models/ && tar xf ch_PP-OCRv3_det_distill_train.tar && cd ../
+ fi
+ if [ ${model_name} == "en_table_structure" ];then
+ wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/dygraph_v2.1/table/en_ppocr_mobile_v2.0_table_structure_train.tar --no-check-certificate
+ cd ./pretrain_models/ && tar xf en_ppocr_mobile_v2.0_table_structure_train.tar && cd ../
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_rec_infer.tar --no-check-certificate
+ cd ./inference/ && tar xf en_ppocr_mobile_v2.0_table_det_infer.tar && tar xf en_ppocr_mobile_v2.0_table_rec_infer.tar && cd ../
+ fi
cd ./pretrain_models/ && tar xf det_mv3_db_v2.0_train.tar && cd ../
rm -rf ./train_data/icdar2015
rm -rf ./train_data/ic15_data
+ rm -rf ./train_data/pubtabnet
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/icdar2015_lite.tar --no-check-certificate
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ic15_data.tar --no-check-certificate
+ wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dataset/pubtabnet.tar --no-check-certificate
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
wget -nc -P ./deploy/slim/prune https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/sen.pickle --no-check-certificate
- cd ./train_data/ && tar xf icdar2015_lite.tar && tar xf ic15_data.tar
+ cd ./train_data/ && tar xf icdar2015_lite.tar && tar xf ic15_data.tar && tar xf pubtabnet.tar
ln -s ./icdar2015_lite ./icdar2015
+ wget -nc -P ./ic15_data/ https://paddleocr.bj.bcebos.com/dataset/rec_gt_train_lite.txt --no-check-certificate
+ wget -nc -P ./ic15_data/ https://paddleocr.bj.bcebos.com/dataset/rec_gt_test_lite.txt --no-check-certificate
cd ../
cd ./inference && tar xf rec_inference.tar && cd ../
- if [ ${model_name} == "ch_PPOCRv2_det" ] || [ ${model_name} == "ch_PPOCRv2_det_PACT" ]; then
+ if [ ${model_name} == "ch_PP-OCRv2_det" ] || [ ${model_name} == "ch_PP-OCRv2_det_PACT" ]; then
wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_train.tar --no-check-certificate
cd ./pretrain_models/ && tar xf ch_ppocr_server_v2.0_det_train.tar && cd ../
fi
- if [ ${model_name} == "ch_PPOCRv2_rec" ] || [ ${model_name} == "ch_PPOCRv2_rec_PACT" ]; then
+ if [ ${model_name} == "ch_PP-OCRv2_rec" ] || [ ${model_name} == "ch_PP-OCRv2_rec_PACT" ]; then
wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_train.tar --no-check-certificate
cd ./pretrain_models/ && tar xf ch_PP-OCRv2_rec_train.tar && cd ../
fi
+ if [ ${model_name} == "ch_PP-OCRv3_rec" ] || [ ${model_name} == "ch_PP-OCRv3_rec_PACT" ]; then
+ wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_train.tar --no-check-certificate
+ cd ./pretrain_models/ && tar xf ch_PP-OCRv3_rec_train.tar && cd ../
+ fi
if [ ${model_name} == "det_r18_db_v2_0" ]; then
wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/pretrained/ResNet18_vd_pretrained.pdparams --no-check-certificate
fi
@@ -79,8 +98,10 @@ if [ ${MODE} = "lite_train_lite_infer" ];then
fi
if [ ${model_name} == "det_r50_vd_sast_icdar15_v2.0" ] || [ ${model_name} == "det_r50_vd_sast_totaltext_v2.0" ]; then
wget -nc -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/ResNet50_vd_ssld_pretrained.pdparams --no-check-certificate
+ wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_sast_icdar15_v2.0_train.tar --no-check-certificate
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/total_text_lite.tar --no-check-certificate
cd ./train_data && tar xf total_text_lite.tar && ln -s total_text_lite total_text && cd ../
+ cd ./pretrain_models && tar xf det_r50_vd_sast_icdar15_v2.0_train.tar && cd ../
fi
if [ ${model_name} == "det_mv3_db_v2_0" ]; then
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_mv3_db_v2.0_train.tar --no-check-certificate
@@ -104,13 +125,22 @@ elif [ ${MODE} = "whole_train_whole_infer" ];then
wget -nc -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/MobileNetV3_large_x0_5_pretrained.pdparams --no-check-certificate
rm -rf ./train_data/icdar2015
rm -rf ./train_data/ic15_data
+ rm -rf ./train_data/pubtabnet
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/icdar2015.tar --no-check-certificate
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ic15_data.tar --no-check-certificate
- cd ./train_data/ && tar xf icdar2015.tar && tar xf ic15_data.tar && cd ../
- if [ ${model_name} == "ch_PPOCRv2_det" ]; then
+ wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dataset/pubtabnet.tar --no-check-certificate
+ cd ./train_data/ && tar xf icdar2015.tar && tar xf ic15_data.tar && tar xf pubtabnet.tar
+ wget -nc -P ./ic15_data/ https://paddleocr.bj.bcebos.com/dataset/rec_gt_train_lite.txt --no-check-certificate
+ wget -nc -P ./ic15_data/ https://paddleocr.bj.bcebos.com/dataset/rec_gt_test_lite.txt --no-check-certificate
+ cd ../
+ if [ ${model_name} == "ch_PP-OCRv2_det" ]; then
wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_distill_train.tar --no-check-certificate
cd ./pretrain_models/ && tar xf ch_PP-OCRv2_det_distill_train.tar && cd ../
fi
+ if [ ${model_name} == "ch_PP-OCRv3_det" ]; then
+ wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_distill_train.tar --no-check-certificate
+ cd ./pretrain_models/ && tar xf ch_PP-OCRv3_det_distill_train.tar && cd ../
+ fi
if [ ${model_name} == "en_server_pgnetA" ]; then
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/total_text_lite.tar --no-check-certificate
wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/en_server_pgnetA.tar --no-check-certificate
@@ -122,19 +152,41 @@ elif [ ${MODE} = "whole_train_whole_infer" ];then
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/total_text_lite.tar --no-check-certificate
cd ./train_data && tar xf total_text.tar && ln -s total_text_lite total_text && cd ../
fi
+ if [[ ${model_name} =~ "en_table_structure" ]];then
+ wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/dygraph_v2.1/table/en_ppocr_mobile_v2.0_table_structure_train.tar --no-check-certificate
+ cd ./pretrain_models/ && tar xf en_ppocr_mobile_v2.0_table_structure_train.tar && cd ../
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_rec_infer.tar --no-check-certificate
+ cd ./inference/ && tar xf en_ppocr_mobile_v2.0_table_det_infer.tar && tar xf en_ppocr_mobile_v2.0_table_rec_infer.tar && cd ../
+ fi
elif [ ${MODE} = "lite_train_whole_infer" ];then
wget -nc -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/MobileNetV3_large_x0_5_pretrained.pdparams --no-check-certificate
rm -rf ./train_data/icdar2015
rm -rf ./train_data/ic15_data
+ rm -rf ./train_data/pubtabnet
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/icdar2015_infer.tar --no-check-certificate
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ic15_data.tar --no-check-certificate
- cd ./train_data/ && tar xf icdar2015_infer.tar && tar xf ic15_data.tar
+ wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dataset/pubtabnet.tar --no-check-certificate
+ cd ./train_data/ && tar xf icdar2015_infer.tar && tar xf ic15_data.tar && tar xf pubtabnet.tar
ln -s ./icdar2015_infer ./icdar2015
+ wget -nc -P ./ic15_data/ https://paddleocr.bj.bcebos.com/dataset/rec_gt_train_lite.txt --no-check-certificate
+ wget -nc -P ./ic15_data/ https://paddleocr.bj.bcebos.com/dataset/rec_gt_test_lite.txt --no-check-certificate
cd ../
- if [ ${model_name} == "ch_PPOCRv2_det" ]; then
+ if [ ${model_name} == "ch_PP-OCRv2_det" ]; then
wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_distill_train.tar --no-check-certificate
cd ./pretrain_models/ && tar xf ch_PP-OCRv2_det_distill_train.tar && cd ../
fi
+ if [ ${model_name} == "ch_PP-OCRv3_det" ]; then
+ wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_distill_train.tar --no-check-certificate
+ cd ./pretrain_models/ && tar xf ch_PP-OCRv3_det_distill_train.tar && cd ../
+ fi
+ if [[ ${model_name} =~ "en_table_structure" ]];then
+ wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/dygraph_v2.1/table/en_ppocr_mobile_v2.0_table_structure_train.tar --no-check-certificate
+ cd ./pretrain_models/ && tar xf en_ppocr_mobile_v2.0_table_structure_train.tar && cd ../
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_rec_infer.tar --no-check-certificate
+ cd ./inference/ && tar xf en_ppocr_mobile_v2.0_table_det_infer.tar && tar xf en_ppocr_mobile_v2.0_table_rec_infer.tar && cd ../
+ fi
elif [ ${MODE} = "whole_infer" ];then
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
@@ -169,17 +221,42 @@ elif [ ${MODE} = "whole_infer" ];then
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar --no-check-certificate
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
fi
- if [[ ${model_name} =~ "ch_PPOCRv2_det" ]]; then
+ if [[ ${model_name} =~ "ch_PP-OCRv2" ]]; then
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv2_det_infer.tar && tar xf ch_PP-OCRv2_rec_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ fi
+ if [[ ${model_name} =~ "ch_PP-OCRv3" ]]; then
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv3_det_infer.tar && tar xf ch_PP-OCRv3_rec_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ fi
+ if [[ ${model_name} =~ "ch_PP-OCRv2_det" ]]; then
eval_model_name="ch_PP-OCRv2_det_infer"
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar --no-check-certificate
cd ./inference && tar xf ${eval_model_name}.tar && tar xf ch_det_data_50.tar && cd ../
fi
- if [[ ${model_name} =~ "PPOCRv2_ocr_rec" ]]; then
+ if [[ ${model_name} =~ "ch_PP-OCRv3_det" ]]; then
+ eval_model_name="ch_PP-OCRv3_det_infer"
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar --no-check-certificate
+ cd ./inference && tar xf ${eval_model_name}.tar && tar xf ch_det_data_50.tar && cd ../
+ fi
+ if [[ ${model_name} =~ "ch_PP-OCRv2_rec" ]]; then
eval_model_name="ch_PP-OCRv2_rec_infer"
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar --no-check-certificate
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_slim_quant_infer.tar --no-check-certificate
cd ./inference && tar xf ${eval_model_name}.tar && tar xf ch_PP-OCRv2_rec_slim_quant_infer.tar && cd ../
fi
+ if [[ ${model_name} =~ "ch_PP-OCRv3_rec" ]]; then
+ eval_model_name="ch_PP-OCRv3_rec_infer"
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tar --no-check-certificate
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_slim_infer.tar --no-check-certificate
+ cd ./inference && tar xf ${eval_model_name}.tar && tar xf ch_PP-OCRv3_rec_slim_infer.tar && cd ../
+ fi
+ if [[ ${model_name} == "ch_PP-OCRv3_rec_PACT" ]]; then
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_slim_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv3_rec_slim_infer.tar && cd ../
+ fi
if [ ${model_name} == "en_server_pgnetA" ]; then
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/en_server_pgnetA.tar --no-check-certificate
cd ./inference && tar xf en_server_pgnetA.tar && tar xf ch_det_data_50.tar && cd ../
@@ -269,9 +346,15 @@ elif [ ${MODE} = "whole_infer" ];then
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_east_v2.0_train.tar --no-check-certificate
cd ./inference/ && tar xf det_r50_vd_east_v2.0_train.tar & cd ../
fi
+ if [[ ${model_name} =~ "en_table_structure" ]];then
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_structure_infer.tar --no-check-certificate
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_rec_infer.tar --no-check-certificate
+ cd ./inference/ && tar xf en_ppocr_mobile_v2.0_table_structure_infer.tar && tar xf en_ppocr_mobile_v2.0_table_det_infer.tar && tar xf en_ppocr_mobile_v2.0_table_rec_infer.tar && cd ../
+ fi
fi
-if [ ${MODE} = "klquant_whole_infer" ]; then
+if [[ ${model_name} =~ "KL" ]]; then
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/icdar2015_lite.tar --no-check-certificate
cd ./train_data/ && tar xf icdar2015_lite.tar && rm -rf ./icdar2015 && ln -s ./icdar2015_lite ./icdar2015 && cd ../
if [ ${model_name} = "ch_ppocr_mobile_v2.0_det_KL" ]; then
@@ -279,41 +362,151 @@ if [ ${MODE} = "klquant_whole_infer" ]; then
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
cd ./inference && tar xf ch_ppocr_mobile_v2.0_det_infer.tar && tar xf ch_det_data_50.tar && cd ../
fi
- if [ ${model_name} = "PPOCRv2_ocr_rec_kl" ]; then
+ if [ ${model_name} = "ch_PP-OCRv2_rec_KL" ]; then
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar --no-check-certificate
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ic15_data.tar --no-check-certificate
cd ./train_data/ && tar xf ic15_data.tar && cd ../
cd ./inference && tar xf rec_inference.tar && tar xf ch_PP-OCRv2_rec_infer.tar && cd ../
fi
- if [ ${model_name} = "PPOCRv2_ocr_det_kl" ]; then
+ if [ ${model_name} = "ch_PP-OCRv3_rec_KL" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tar --no-check-certificate
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
+ wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ic15_data.tar --no-check-certificate
+ cd ./train_data/ && tar xf ic15_data.tar
+ wget -nc -P ./ic15_data/ https://paddleocr.bj.bcebos.com/dataset/rec_gt_train_lite.txt --no-check-certificate
+ wget -nc -P ./ic15_data/ https://paddleocr.bj.bcebos.com/dataset/rec_gt_test_lite.txt --no-check-certificate
+ cd ../
+ cd ./inference && tar xf rec_inference.tar && tar xf ch_PP-OCRv3_rec_infer.tar && cd ../
+ fi
+ if [ ${model_name} = "ch_PP-OCRv2_det_KL" ]; then
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar --no-check-certificate
cd ./inference && tar xf ch_PP-OCRv2_det_infer.tar && tar xf ch_det_data_50.tar && cd ../
fi
+ if [ ${model_name} = "ch_PP-OCRv3_det_KL" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv3_det_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ fi
if [ ${model_name} = "ch_ppocr_mobile_v2.0_rec_KL" ]; then
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_infer.tar --no-check-certificate
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ic15_data.tar --no-check-certificate
cd ./train_data/ && tar xf ic15_data.tar && cd ../
cd ./inference && tar xf ch_ppocr_mobile_v2.0_rec_infer.tar && tar xf rec_inference.tar && cd ../
- fi
+ fi
+ if [ ${model_name} = "en_table_structure_KL" ];then
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_structure_infer.tar --no-check-certificate
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_rec_infer.tar --no-check-certificate
+ wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dataset/pubtabnet.tar --no-check-certificate
+ cd ./inference/ && tar xf en_ppocr_mobile_v2.0_table_structure_infer.tar && tar xf en_ppocr_mobile_v2.0_table_det_infer.tar && tar xf en_ppocr_mobile_v2.0_table_rec_infer.tar && cd ../
+ cd ./train_data/ && tar xf pubtabnet.tar && cd ../
+ fi
fi
if [ ${MODE} = "cpp_infer" ];then
- if [ ${model_name} = "ocr_det" ]; then
+ if [ ${model_name} = "ch_ppocr_mobile_v2.0_det" ]; then
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_infer.tar --no-check-certificate
cd ./inference && tar xf ch_ppocr_mobile_v2.0_det_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ elif [ ${model_name} = "ch_ppocr_mobile_v2.0_det_KL" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_ppocr_mobile_v2.0_det_klquant_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_ppocr_mobile_v2.0_det_klquant_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ elif [ ${model_name} = "ch_ppocr_mobile_v2.0_det_PACT" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_ppocr_mobile_v2.0_det_pact_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_ppocr_mobile_v2.0_det_pact_infer.tar && tar xf ch_det_data_50.tar && cd ../
elif [ ${model_name} = "ch_ppocr_mobile_v2.0_rec" ]; then
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_infer.tar --no-check-certificate
cd ./inference && tar xf ch_ppocr_mobile_v2.0_rec_infer.tar && tar xf rec_inference.tar && cd ../
- elif [ ${model_name} = "ocr_system" ]; then
+ elif [ ${model_name} = "ch_ppocr_mobile_v2.0_rec_KL" ]; then
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_ppocr_mobile_v2.0_rec_klquant_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_ppocr_mobile_v2.0_rec_klquant_infer.tar && tar xf rec_inference.tar && cd ../
+ elif [ ${model_name} = "ch_ppocr_mobile_v2.0_rec_PACT" ]; then
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_ppocr_mobile_v2.0_rec_pact_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_ppocr_mobile_v2.0_rec_pact_infer.tar && tar xf rec_inference.tar && cd ../
+ elif [ ${model_name} = "ch_ppocr_server_v2.0_det" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_ppocr_server_v2.0_det_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ elif [ ${model_name} = "ch_ppocr_server_v2.0_rec" ]; then
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_ppocr_server_v2.0_rec_infer.tar && tar xf rec_inference.tar && cd ../
+ elif [ ${model_name} = "ch_PP-OCRv2_det" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv2_det_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ elif [ ${model_name} = "ch_PP-OCRv2_det_KL" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv2_det_klquant_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv2_det_klquant_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ elif [ ${model_name} = "ch_PP-OCRv2_det_PACT" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv2_det_pact_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv2_det_pact_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ elif [ ${model_name} = "ch_PP-OCRv2_rec" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv2_rec_infer.tar && tar xf rec_inference.tar && cd ../
+ elif [ ${model_name} = "ch_PP-OCRv2_rec_KL" ]; then
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv2_rec_klquant_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv2_rec_klquant_infer.tar && tar xf rec_inference.tar && cd ../
+ elif [ ${model_name} = "ch_PP-OCRv2_rec_PACT" ]; then
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv2_rec_pact_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv2_rec_pact_infer.tar && tar xf rec_inference.tar && cd ../
+ elif [ ${model_name} = "ch_PP-OCRv3_det" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv3_det_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ elif [ ${model_name} = "ch_PP-OCRv3_det_KL" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv3_det_klquant_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv3_det_klquant_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ elif [ ${model_name} = "ch_PP-OCRv3_det_PACT" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv3_det_pact_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv3_det_pact_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ elif [ ${model_name} = "ch_PP-OCRv3_rec" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv3_rec_infer.tar && tar xf rec_inference.tar && cd ../
+ elif [ ${model_name} = "ch_PP-OCRv3_rec_KL" ]; then
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv3_rec_klquant_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv3_rec_klquant_infer.tar && tar xf rec_inference.tar && cd ../
+ elif [ ${model_name} = "ch_PP-OCRv3_rec_PACT" ]; then
+ wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv3_rec_pact_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv3_rec_pact_infer.tar && tar xf rec_inference.tar && cd ../
+ elif [ ${model_name} = "ch_ppocr_mobile_v2.0" ]; then
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_infer.tar --no-check-certificate
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_infer.tar --no-check-certificate
cd ./inference && tar xf ch_ppocr_mobile_v2.0_det_infer.tar && tar xf ch_ppocr_mobile_v2.0_rec_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ elif [ ${model_name} = "ch_ppocr_server_v2.0" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_ppocr_server_v2.0_det_infer.tar && tar xf ch_ppocr_server_v2.0_rec_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ elif [ ${model_name} = "ch_PP-OCRv2" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv2_det_infer.tar && tar xf ch_PP-OCRv2_rec_infer.tar && tar xf ch_det_data_50.tar && cd ../
+ elif [ ${model_name} = "ch_PP-OCRv3" ]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv3_det_infer.tar && tar xf ch_PP-OCRv3_rec_infer.tar && tar xf ch_det_data_50.tar && cd ../
fi
fi
@@ -323,29 +516,84 @@ if [ ${MODE} = "serving_infer" ];then
IFS='|'
array=(${python_name_list})
python_name=${array[0]}
- ${python_name} -m pip install paddle-serving-server-gpu==0.8.3.post101
- ${python_name} -m pip install paddle_serving_client==0.8.3
- ${python_name} -m pip install paddle-serving-app==0.8.3
- wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_infer.tar
- wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_infer.tar
- wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_infer.tar
- wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_rec_infer.tar
- cd ./inference && tar xf ch_ppocr_mobile_v2.0_det_infer.tar && tar xf ch_ppocr_mobile_v2.0_rec_infer.tar && tar xf ch_ppocr_server_v2.0_rec_infer.tar && tar xf ch_ppocr_server_v2.0_det_infer.tar && cd ../
+ ${python_name} -m pip install paddle-serving-server-gpu
+ ${python_name} -m pip install paddle_serving_client
+ ${python_name} -m pip install paddle-serving-app
+ # wget model
+ if [ ${model_name} == "ch_ppocr_mobile_v2.0_det_KL" ] || [ ${model_name} == "ch_ppocr_mobile_v2.0_rec_KL" ] ; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_ppocr_mobile_v2.0_det_klquant_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_ppocr_mobile_v2.0_rec_klquant_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_ppocr_mobile_v2.0_det_klquant_infer.tar && tar xf ch_ppocr_mobile_v2.0_rec_klquant_infer.tar && cd ../
+ elif [ ${model_name} == "ch_PP-OCRv2_det_KL" ] || [ ${model_name} == "ch_PP-OCRv2_rec_KL" ] ; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv2_det_klquant_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv2_rec_klquant_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv2_det_klquant_infer.tar && tar xf ch_PP-OCRv2_rec_klquant_infer.tar && cd ../
+ elif [ ${model_name} == "ch_PP-OCRv3_det_KL" ] || [ ${model_name} == "ch_PP-OCRv3_rec_KL" ] ; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv3_det_klquant_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv3_rec_klquant_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv3_det_klquant_infer.tar && tar xf ch_PP-OCRv3_rec_klquant_infer.tar && cd ../
+ elif [ ${model_name} == "ch_ppocr_mobile_v2.0_det_PACT" ] || [ ${model_name} == "ch_ppocr_mobile_v2.0_rec_PACT" ] ; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_ppocr_mobile_v2.0_det_pact_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_ppocr_mobile_v2.0_rec_pact_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_ppocr_mobile_v2.0_det_pact_infer.tar && tar xf ch_ppocr_mobile_v2.0_rec_pact_infer.tar && cd ../
+ elif [ ${model_name} == "ch_PP-OCRv2_det_PACT" ] || [ ${model_name} == "ch_PP-OCRv2_rec_PACT" ] ; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv2_det_pact_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv2_rec_pact_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv2_det_pact_infer.tar && tar xf ch_PP-OCRv2_rec_pact_infer.tar && cd ../
+ elif [ ${model_name} == "ch_PP-OCRv3_det_PACT" ] || [ ${model_name} == "ch_PP-OCRv3_rec_PACT" ] ; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv3_det_pact_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/tipc_fake_model/ch_PP-OCRv3_rec_pact_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv3_det_pact_infer.tar && tar xf ch_PP-OCRv3_rec_pact_infer.tar && cd ../
+ elif [[ ${model_name} =~ "ch_ppocr_mobile_v2.0" ]]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_ppocr_mobile_v2.0_det_infer.tar && tar xf ch_ppocr_mobile_v2.0_rec_infer.tar && cd ../
+ elif [[ ${model_name} =~ "ch_ppocr_server_v2.0" ]]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_ppocr_server_v2.0_det_infer.tar && tar xf ch_ppocr_server_v2.0_rec_infer.tar && cd ../
+ elif [[ ${model_name} =~ "ch_PP-OCRv2" ]]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv2_det_infer.tar && tar xf ch_PP-OCRv2_rec_infer.tar && cd ../
+ elif [[ ${model_name} =~ "ch_PP-OCRv3" ]]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv3_det_infer.tar && tar xf ch_PP-OCRv3_rec_infer.tar && cd ../
+ fi
+ # wget data
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
+ cd ./inference && tar xf ch_det_data_50.tar && tar xf rec_inference.tar && cd ../
fi
if [ ${MODE} = "paddle2onnx_infer" ];then
# prepare serving env
python_name=$(func_parser_value "${lines[2]}")
- ${python_name} -m pip install install paddle2onnx
- ${python_name} -m pip install onnxruntime==1.4.0
+ ${python_name} -m pip install paddle2onnx
+ ${python_name} -m pip install onnxruntime
# wget model
- wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_infer.tar
- wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_infer.tar
- wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_infer.tar
- wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_rec_infer.tar
+ if [[ ${model_name} =~ "ch_ppocr_mobile_v2.0" ]]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_ppocr_mobile_v2.0_det_infer.tar && tar xf ch_ppocr_mobile_v2.0_rec_infer.tar && cd ../
+ elif [[ ${model_name} =~ "ch_ppocr_server_v2.0" ]]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_ppocr_server_v2.0_det_infer.tar && tar xf ch_ppocr_server_v2.0_rec_infer.tar && cd ../
+ elif [[ ${model_name} =~ "ch_PP-OCRv2" ]]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv2_det_infer.tar && tar xf ch_PP-OCRv2_rec_infer.tar && cd ../
+ elif [[ ${model_name} =~ "ch_PP-OCRv3" ]]; then
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar --no-check-certificate
+ wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tar --no-check-certificate
+ cd ./inference && tar xf ch_PP-OCRv3_det_infer.tar && tar xf ch_PP-OCRv3_rec_infer.tar && cd ../
+ fi
+
# wget data
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar
- cd ./inference && tar xf ch_ppocr_mobile_v2.0_det_infer.tar && tar xf ch_ppocr_mobile_v2.0_rec_infer.tar && tar xf ch_ppocr_server_v2.0_rec_infer.tar && tar xf ch_ppocr_server_v2.0_det_infer.tar && tar xf ch_det_data_50.tar && tar xf rec_inference.tar && cd ../
+ cd ./inference && tar xf ch_det_data_50.tar && tar xf rec_inference.tar && cd ../
fi
diff --git a/test_tipc/prepare_lite_cpp.sh b/test_tipc/prepare_lite_cpp.sh
index 94af43c88d..9148cb5dd7 100644
--- a/test_tipc/prepare_lite_cpp.sh
+++ b/test_tipc/prepare_lite_cpp.sh
@@ -51,6 +51,8 @@ for model in ${lite_model_list[*]}; do
inference_model_url=https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/${model}.tar
elif [[ $model =~ "v2.0" ]]; then
inference_model_url=https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/${model}.tar
+ elif [[ $model =~ "PP-OCRv3" ]]; then
+ inference_model_url=https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/${model}.tar
else
echo "Model is wrong, please check."
exit 3
diff --git a/test_tipc/readme.md b/test_tipc/readme.md
index 8110f0073b..effb2f168b 100644
--- a/test_tipc/readme.md
+++ b/test_tipc/readme.md
@@ -138,6 +138,7 @@ bash test_tipc/test_train_inference_python.sh ./test_tipc/configs/ch_ppocr_mobil
## 4. 开始测试
各功能测试中涉及混合精度、裁剪、量化等训练相关，及mkldnn、Tensorrt等多种预测相关参数配置，请点击下方相应链接了解更多细节和使用教程：
- [test_train_inference_python 使用](docs/test_train_inference_python.md) ：测试基于Python的模型训练、评估、推理等基本功能，包括裁剪、量化、蒸馏。
+- [test_train_fleet_inference_python 使用](./docs/test_train_fleet_inference_python.md)：测试基于Python的多机多卡训练与推理等基本功能。
- [test_inference_cpp 使用](docs/test_inference_cpp.md)：测试基于C++的模型推理。
- [test_serving 使用](docs/test_serving.md)：测试基于Paddle Serving的服务化部署功能。
- [test_lite_arm_cpp 使用](docs/test_lite_arm_cpp.md)：测试基于Paddle-Lite的ARM CPU端c++预测部署功能。
diff --git a/test_tipc/test_inference_cpp.sh b/test_tipc/test_inference_cpp.sh
index 9885e39372..c0c7c18a38 100644
--- a/test_tipc/test_inference_cpp.sh
+++ b/test_tipc/test_inference_cpp.sh
@@ -43,7 +43,7 @@ cpp_cls_value=$(func_parser_value "${lines[18]}")
cpp_use_angle_cls_key=$(func_parser_key "${lines[19]}")
cpp_use_angle_cls_value=$(func_parser_value "${lines[19]}")
-LOG_PATH="./test_tipc/output"
+LOG_PATH="./test_tipc/output/${model_name}/cpp_infer"
mkdir -p ${LOG_PATH}
status_log="${LOG_PATH}/results_cpp.log"
@@ -84,7 +84,7 @@ function func_cpp_inference(){
eval $command
last_status=${PIPESTATUS[0]}
eval "cat ${_save_log_path}"
- status_check $last_status "${command}" "${status_log}"
+ status_check $last_status "${command}" "${status_log}" "${model_name}"
done
done
done
@@ -117,7 +117,7 @@ function func_cpp_inference(){
eval $command
last_status=${PIPESTATUS[0]}
eval "cat ${_save_log_path}"
- status_check $last_status "${command}" "${status_log}"
+ status_check $last_status "${command}" "${status_log}" "${model_name}"
done
done
@@ -178,7 +178,23 @@ if [ ${use_opencv} = "True" ]; then
else
OPENCV_DIR=''
fi
-LIB_DIR=$(pwd)/Paddle/build/paddle_inference_install_dir/
+if [ -d "paddle_inference/" ] ;then
+ echo "################### download paddle inference skipped ###################"
+else
+ echo "################### download paddle inference ###################"
+ PADDLEInfer=$3
+ if [ "" = "$PADDLEInfer" ];then
+ wget -nc https://paddle-inference-lib.bj.bcebos.com/2.3.0/cxx_c/Linux/GPU/x86-64_gcc8.2_avx_mkl_cuda10.1_cudnn7.6.5_trt6.0.1.5/paddle_inference.tgz --no-check-certificate
+ else
+ wget -nc $PADDLEInfer --no-check-certificate
+ fi
+ tar zxf paddle_inference.tgz
+ if [ ! -d "paddle_inference" ]; then
+ ln -s paddle_inference_install_dir paddle_inference
+ fi
+ echo "################### download paddle inference finished ###################"
+fi
+LIB_DIR=$(pwd)/paddle_inference/
CUDA_LIB_DIR=$(dirname `find /usr -name libcudart.so`)
CUDNN_LIB_DIR=$(dirname `find /usr -name libcudnn.so`)
@@ -205,11 +221,10 @@ echo "################### build PaddleOCR demo finished ###################"
# set cuda device
GPUID=$2
if [ ${#GPUID} -le 0 ];then
- env=" "
+ env="export CUDA_VISIBLE_DEVICES=0"
else
env="export CUDA_VISIBLE_DEVICES=${GPUID}"
fi
-set CUDA_VISIBLE_DEVICES
eval $env
diff --git a/test_tipc/test_inference_python.sh b/test_tipc/test_inference_python.sh
index 27276d55b9..2a31a468f0 100644
--- a/test_tipc/test_inference_python.sh
+++ b/test_tipc/test_inference_python.sh
@@ -44,7 +44,7 @@ infer_value1=$(func_parser_value "${lines[17]}")
-LOG_PATH="./test_tipc/output"
+LOG_PATH="./test_tipc/output/${model_name}/${MODE}"
mkdir -p ${LOG_PATH}
status_log="${LOG_PATH}/results_python.log"
@@ -88,7 +88,7 @@ function func_inference(){
eval $command
last_status=${PIPESTATUS[0]}
eval "cat ${_save_log_path}"
- status_check $last_status "${command}" "${status_log}"
+ status_check $last_status "${command}" "${status_log}" "${model_name}"
done
done
done
@@ -113,13 +113,13 @@ function func_inference(){
set_tensorrt=$(func_set_params "${use_trt_key}" "${use_trt}")
set_precision=$(func_set_params "${precision_key}" "${precision}")
set_model_dir=$(func_set_params "${infer_model_key}" "${_model_dir}")
- set_infer_params0=$(func_set_params "${save_log_key}" "${save_log_value}")
+ set_infer_params0=$(func_set_params "${rec_model_key}" "${rec_model_value}")
set_infer_params1=$(func_set_params "${infer_key1}" "${infer_value1}")
command="${_python} ${_script} ${use_gpu_key}=${use_gpu} ${set_tensorrt} ${set_precision} ${set_model_dir} ${set_batchsize} ${set_infer_data} ${set_benchmark} ${set_infer_params1} ${set_infer_params0} > ${_save_log_path} 2>&1 "
eval $command
last_status=${PIPESTATUS[0]}
eval "cat ${_save_log_path}"
- status_check $last_status "${command}" "${status_log}"
+ status_check $last_status "${command}" "${status_log}" "${model_name}"
done
done
@@ -153,7 +153,7 @@ if [ ${MODE} = "whole_infer" ]; then
echo ${infer_run_exports[Count]}
eval $export_cmd
status_export=$?
- status_check $status_export "${export_cmd}" "${status_log}"
+ status_check $status_export "${export_cmd}" "${status_log}" "${model_name}"
else
save_infer_dir=${infer_model}
fi
diff --git a/test_tipc/test_paddle2onnx.sh b/test_tipc/test_paddle2onnx.sh
index 300c61770d..356bc98041 100644
--- a/test_tipc/test_paddle2onnx.sh
+++ b/test_tipc/test_paddle2onnx.sh
@@ -11,7 +11,7 @@ python=$(func_parser_value "${lines[2]}")
# parser params
-dataline=$(awk 'NR==1, NR==12{print}' $FILENAME)
+dataline=$(awk 'NR==1, NR==17{print}' $FILENAME)
IFS=$'\n'
lines=(${dataline})
@@ -19,29 +19,33 @@ lines=(${dataline})
model_name=$(func_parser_value "${lines[1]}")
python=$(func_parser_value "${lines[2]}")
padlle2onnx_cmd=$(func_parser_value "${lines[3]}")
-infer_model_dir_key=$(func_parser_key "${lines[4]}")
-infer_model_dir_value=$(func_parser_value "${lines[4]}")
+det_infer_model_dir_key=$(func_parser_key "${lines[4]}")
+det_infer_model_dir_value=$(func_parser_value "${lines[4]}")
model_filename_key=$(func_parser_key "${lines[5]}")
model_filename_value=$(func_parser_value "${lines[5]}")
params_filename_key=$(func_parser_key "${lines[6]}")
params_filename_value=$(func_parser_value "${lines[6]}")
-save_file_key=$(func_parser_key "${lines[7]}")
-save_file_value=$(func_parser_value "${lines[7]}")
-opset_version_key=$(func_parser_key "${lines[8]}")
-opset_version_value=$(func_parser_value "${lines[8]}")
-enable_onnx_checker_key=$(func_parser_key "${lines[9]}")
-enable_onnx_checker_value=$(func_parser_value "${lines[9]}")
+det_save_file_key=$(func_parser_key "${lines[7]}")
+det_save_file_value=$(func_parser_value "${lines[7]}")
+rec_infer_model_dir_key=$(func_parser_key "${lines[8]}")
+rec_infer_model_dir_value=$(func_parser_value "${lines[8]}")
+rec_save_file_key=$(func_parser_key "${lines[9]}")
+rec_save_file_value=$(func_parser_value "${lines[9]}")
+opset_version_key=$(func_parser_key "${lines[10]}")
+opset_version_value=$(func_parser_value "${lines[10]}")
+enable_onnx_checker_key=$(func_parser_key "${lines[11]}")
+enable_onnx_checker_value=$(func_parser_value "${lines[11]}")
# parser onnx inference
-inference_py=$(func_parser_value "${lines[10]}")
-use_gpu_key=$(func_parser_key "${lines[11]}")
-use_gpu_value=$(func_parser_value "${lines[11]}")
-det_model_key=$(func_parser_key "${lines[12]}")
-image_dir_key=$(func_parser_key "${lines[13]}")
-image_dir_value=$(func_parser_value "${lines[13]}")
+inference_py=$(func_parser_value "${lines[12]}")
+use_gpu_key=$(func_parser_key "${lines[13]}")
+use_gpu_list=$(func_parser_value "${lines[13]}")
+det_model_key=$(func_parser_key "${lines[14]}")
+rec_model_key=$(func_parser_key "${lines[15]}")
+image_dir_key=$(func_parser_key "${lines[16]}")
+image_dir_value=$(func_parser_value "${lines[16]}")
-
-LOG_PATH="./test_tipc/output"
-mkdir -p ./test_tipc/output
+LOG_PATH="./test_tipc/output/${model_name}/paddle2onnx"
+mkdir -p ${LOG_PATH}
status_log="${LOG_PATH}/results_paddle2onnx.log"
@@ -50,24 +54,103 @@ function func_paddle2onnx(){
_script=$1
# paddle2onnx
- _save_log_path="${LOG_PATH}/paddle2onnx_infer_cpu.log"
- set_dirname=$(func_set_params "${infer_model_dir_key}" "${infer_model_dir_value}")
- set_model_filename=$(func_set_params "${model_filename_key}" "${model_filename_value}")
- set_params_filename=$(func_set_params "${params_filename_key}" "${params_filename_value}")
- set_save_model=$(func_set_params "${save_file_key}" "${save_file_value}")
- set_opset_version=$(func_set_params "${opset_version_key}" "${opset_version_value}")
- set_enable_onnx_checker=$(func_set_params "${enable_onnx_checker_key}" "${enable_onnx_checker_value}")
- trans_model_cmd="${padlle2onnx_cmd} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_save_model} ${set_opset_version} ${set_enable_onnx_checker}"
- eval $trans_model_cmd
- last_status=${PIPESTATUS[0]}
- status_check $last_status "${trans_model_cmd}" "${status_log}"
+ if [ ${model_name} = "ch_PP-OCRv2" ] || [ ${model_name} = "ch_PP-OCRv3" ] || [ ${model_name} = "ch_ppocr_mobile_v2.0" ] || [ ${model_name} = "ch_ppocr_server_v2.0" ]; then
+ # trans det
+ set_dirname=$(func_set_params "--model_dir" "${det_infer_model_dir_value}")
+ set_model_filename=$(func_set_params "${model_filename_key}" "${model_filename_value}")
+ set_params_filename=$(func_set_params "${params_filename_key}" "${params_filename_value}")
+ set_save_model=$(func_set_params "--save_file" "${det_save_file_value}")
+ set_opset_version=$(func_set_params "${opset_version_key}" "${opset_version_value}")
+ set_enable_onnx_checker=$(func_set_params "${enable_onnx_checker_key}" "${enable_onnx_checker_value}")
+ trans_det_log="${LOG_PATH}/trans_model_det.log"
+ trans_model_cmd="${padlle2onnx_cmd} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_save_model} ${set_opset_version} ${set_enable_onnx_checker} > ${trans_det_log} 2>&1 "
+ eval $trans_model_cmd
+ last_status=${PIPESTATUS[0]}
+ status_check $last_status "${trans_model_cmd}" "${status_log}" "${model_name}"
+ # trans rec
+ set_dirname=$(func_set_params "--model_dir" "${rec_infer_model_dir_value}")
+ set_model_filename=$(func_set_params "${model_filename_key}" "${model_filename_value}")
+ set_params_filename=$(func_set_params "${params_filename_key}" "${params_filename_value}")
+ set_save_model=$(func_set_params "--save_file" "${rec_save_file_value}")
+ set_opset_version=$(func_set_params "${opset_version_key}" "${opset_version_value}")
+ set_enable_onnx_checker=$(func_set_params "${enable_onnx_checker_key}" "${enable_onnx_checker_value}")
+ trans_rec_log="${LOG_PATH}/trans_model_rec.log"
+ trans_model_cmd="${padlle2onnx_cmd} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_save_model} ${set_opset_version} ${set_enable_onnx_checker} > ${trans_rec_log} 2>&1 "
+ eval $trans_model_cmd
+ last_status=${PIPESTATUS[0]}
+ status_check $last_status "${trans_model_cmd}" "${status_log}" "${model_name}"
+ elif [[ ${model_name} =~ "det" ]]; then
+ # trans det
+ set_dirname=$(func_set_params "--model_dir" "${det_infer_model_dir_value}")
+ set_model_filename=$(func_set_params "${model_filename_key}" "${model_filename_value}")
+ set_params_filename=$(func_set_params "${params_filename_key}" "${params_filename_value}")
+ set_save_model=$(func_set_params "--save_file" "${det_save_file_value}")
+ set_opset_version=$(func_set_params "${opset_version_key}" "${opset_version_value}")
+ set_enable_onnx_checker=$(func_set_params "${enable_onnx_checker_key}" "${enable_onnx_checker_value}")
+ trans_det_log="${LOG_PATH}/trans_model_det.log"
+ trans_model_cmd="${padlle2onnx_cmd} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_save_model} ${set_opset_version} ${set_enable_onnx_checker} > ${trans_det_log} 2>&1 "
+ eval $trans_model_cmd
+ last_status=${PIPESTATUS[0]}
+ status_check $last_status "${trans_model_cmd}" "${status_log}" "${model_name}"
+ elif [[ ${model_name} =~ "rec" ]]; then
+ # trans rec
+ set_dirname=$(func_set_params "--model_dir" "${rec_infer_model_dir_value}")
+ set_model_filename=$(func_set_params "${model_filename_key}" "${model_filename_value}")
+ set_params_filename=$(func_set_params "${params_filename_key}" "${params_filename_value}")
+ set_save_model=$(func_set_params "--save_file" "${rec_save_file_value}")
+ set_opset_version=$(func_set_params "${opset_version_key}" "${opset_version_value}")
+ set_enable_onnx_checker=$(func_set_params "${enable_onnx_checker_key}" "${enable_onnx_checker_value}")
+ trans_rec_log="${LOG_PATH}/trans_model_rec.log"
+ trans_model_cmd="${padlle2onnx_cmd} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_save_model} ${set_opset_version} ${set_enable_onnx_checker} > ${trans_rec_log} 2>&1 "
+ eval $trans_model_cmd
+ last_status=${PIPESTATUS[0]}
+ status_check $last_status "${trans_model_cmd}" "${status_log}" "${model_name}"
+ fi
+
# python inference
- set_gpu=$(func_set_params "${use_gpu_key}" "${use_gpu_value}")
- set_model_dir=$(func_set_params "${det_model_key}" "${save_file_value}")
- set_img_dir=$(func_set_params "${image_dir_key}" "${image_dir_value}")
- infer_model_cmd="${python} ${inference_py} ${set_gpu} ${set_img_dir} ${set_model_dir} --use_onnx=True > ${_save_log_path} 2>&1 "
- eval $infer_model_cmd
- status_check $last_status "${infer_model_cmd}" "${status_log}"
+ for use_gpu in ${use_gpu_list[*]}; do
+ if [ ${use_gpu} = "False" ] || [ ${use_gpu} = "cpu" ]; then
+ _save_log_path="${LOG_PATH}/paddle2onnx_infer_cpu.log"
+ set_gpu=$(func_set_params "${use_gpu_key}" "${use_gpu}")
+ set_img_dir=$(func_set_params "${image_dir_key}" "${image_dir_value}")
+ if [ ${model_name} = "ch_PP-OCRv2" ] || [ ${model_name} = "ch_PP-OCRv3" ] || [ ${model_name} = "ch_ppocr_mobile_v2.0" ] || [ ${model_name} = "ch_ppocr_server_v2.0" ]; then
+ set_det_model_dir=$(func_set_params "${det_model_key}" "${det_save_file_value}")
+ set_rec_model_dir=$(func_set_params "${rec_model_key}" "${rec_save_file_value}")
+ infer_model_cmd="${python} ${inference_py} ${set_gpu} ${set_img_dir} ${set_det_model_dir} ${set_rec_model_dir} --use_onnx=True > ${_save_log_path} 2>&1 "
+ elif [[ ${model_name} =~ "det" ]]; then
+ set_det_model_dir=$(func_set_params "${det_model_key}" "${det_save_file_value}")
+ infer_model_cmd="${python} ${inference_py} ${set_gpu} ${set_img_dir} ${set_det_model_dir} --use_onnx=True > ${_save_log_path} 2>&1 "
+ elif [[ ${model_name} =~ "rec" ]]; then
+ set_rec_model_dir=$(func_set_params "${rec_model_key}" "${rec_save_file_value}")
+ infer_model_cmd="${python} ${inference_py} ${set_gpu} ${set_img_dir} ${set_rec_model_dir} --use_onnx=True > ${_save_log_path} 2>&1 "
+ fi
+ eval $infer_model_cmd
+ last_status=${PIPESTATUS[0]}
+ eval "cat ${_save_log_path}"
+ status_check $last_status "${infer_model_cmd}" "${status_log}" "${model_name}"
+ elif [ ${use_gpu} = "True" ] || [ ${use_gpu} = "gpu" ]; then
+ _save_log_path="${LOG_PATH}/paddle2onnx_infer_gpu.log"
+ set_gpu=$(func_set_params "${use_gpu_key}" "${use_gpu}")
+ set_img_dir=$(func_set_params "${image_dir_key}" "${image_dir_value}")
+ if [ ${model_name} = "ch_PP-OCRv2" ] || [ ${model_name} = "ch_PP-OCRv3" ] || [ ${model_name} = "ch_ppocr_mobile_v2.0" ] || [ ${model_name} = "ch_ppocr_server_v2.0" ]; then
+ set_det_model_dir=$(func_set_params "${det_model_key}" "${det_save_file_value}")
+ set_rec_model_dir=$(func_set_params "${rec_model_key}" "${rec_save_file_value}")
+ infer_model_cmd="${python} ${inference_py} ${set_gpu} ${set_img_dir} ${set_det_model_dir} ${set_rec_model_dir} --use_onnx=True > ${_save_log_path} 2>&1 "
+ elif [[ ${model_name} =~ "det" ]]; then
+ set_det_model_dir=$(func_set_params "${det_model_key}" "${det_save_file_value}")
+ infer_model_cmd="${python} ${inference_py} ${set_gpu} ${set_img_dir} ${set_det_model_dir} --use_onnx=True > ${_save_log_path} 2>&1 "
+ elif [[ ${model_name} =~ "rec" ]]; then
+ set_rec_model_dir=$(func_set_params "${rec_model_key}" "${rec_save_file_value}")
+ infer_model_cmd="${python} ${inference_py} ${set_gpu} ${set_img_dir} ${set_rec_model_dir} --use_onnx=True > ${_save_log_path} 2>&1 "
+ fi
+ eval $infer_model_cmd
+ last_status=${PIPESTATUS[0]}
+ eval "cat ${_save_log_path}"
+ status_check $last_status "${infer_model_cmd}" "${status_log}" "${model_name}"
+ else
+ echo "Does not support hardware other than CPU and GPU Currently!"
+ fi
+ done
}
diff --git a/test_tipc/test_ptq_inference_python.sh b/test_tipc/test_ptq_inference_python.sh
new file mode 100644
index 0000000000..c1aa3daa6c
--- /dev/null
+++ b/test_tipc/test_ptq_inference_python.sh
@@ -0,0 +1,158 @@
+#!/bin/bash
+source test_tipc/common_func.sh
+
+FILENAME=$1
+# MODE be one of [''whole_infer']
+MODE=$2
+
+IFS=$'\n'
+# parser klquant_infer params
+
+dataline=$(awk 'NR==1, NR==17{print}' $FILENAME)
+lines=(${dataline})
+model_name=$(func_parser_value "${lines[1]}")
+python=$(func_parser_value "${lines[2]}")
+export_weight=$(func_parser_key "${lines[3]}")
+save_infer_key=$(func_parser_key "${lines[4]}")
+# parser inference model
+infer_model_dir_list=$(func_parser_value "${lines[5]}")
+infer_export_list=$(func_parser_value "${lines[6]}")
+infer_is_quant=$(func_parser_value "${lines[7]}")
+# parser inference
+inference_py=$(func_parser_value "${lines[8]}")
+use_gpu_key=$(func_parser_key "${lines[9]}")
+use_gpu_list=$(func_parser_value "${lines[9]}")
+use_mkldnn_key=$(func_parser_key "${lines[10]}")
+use_mkldnn_list=$(func_parser_value "${lines[10]}")
+cpu_threads_key=$(func_parser_key "${lines[11]}")
+cpu_threads_list=$(func_parser_value "${lines[11]}")
+batch_size_key=$(func_parser_key "${lines[12]}")
+batch_size_list=$(func_parser_value "${lines[12]}")
+use_trt_key=$(func_parser_key "${lines[13]}")
+use_trt_list=$(func_parser_value "${lines[13]}")
+precision_key=$(func_parser_key "${lines[14]}")
+precision_list=$(func_parser_value "${lines[14]}")
+infer_model_key=$(func_parser_key "${lines[15]}")
+image_dir_key=$(func_parser_key "${lines[16]}")
+infer_img_dir=$(func_parser_value "${lines[16]}")
+save_log_key=$(func_parser_key "${lines[17]}")
+save_log_value=$(func_parser_value "${lines[17]}")
+benchmark_key=$(func_parser_key "${lines[18]}")
+benchmark_value=$(func_parser_value "${lines[18]}")
+infer_key1=$(func_parser_key "${lines[19]}")
+infer_value1=$(func_parser_value "${lines[19]}")
+
+
+LOG_PATH="./test_tipc/output/${model_name}/${MODE}"
+mkdir -p ${LOG_PATH}
+status_log="${LOG_PATH}/results_python.log"
+
+
+function func_inference(){
+ IFS='|'
+ _python=$1
+ _script=$2
+ _model_dir=$3
+ _log_path=$4
+ _img_dir=$5
+ _flag_quant=$6
+ # inference
+ for use_gpu in ${use_gpu_list[*]}; do
+ if [ ${use_gpu} = "False" ] || [ ${use_gpu} = "cpu" ]; then
+ for use_mkldnn in ${use_mkldnn_list[*]}; do
+ for threads in ${cpu_threads_list[*]}; do
+ for batch_size in ${batch_size_list[*]}; do
+ for precision in ${precision_list[*]}; do
+ if [ ${use_mkldnn} = "False" ] && [ ${precision} = "fp16" ]; then
+ continue
+ fi # skip when enable fp16 but disable mkldnn
+ if [ ${_flag_quant} = "True" ] && [ ${precision} != "int8" ]; then
+ continue
+ fi # skip when quant model inference but precision is not int8
+ set_precision=$(func_set_params "${precision_key}" "${precision}")
+
+ _save_log_path="${_log_path}/python_infer_cpu_usemkldnn_${use_mkldnn}_threads_${threads}_precision_${precision}_batchsize_${batch_size}.log"
+ set_infer_data=$(func_set_params "${image_dir_key}" "${_img_dir}")
+ set_benchmark=$(func_set_params "${benchmark_key}" "${benchmark_value}")
+ set_batchsize=$(func_set_params "${batch_size_key}" "${batch_size}")
+ set_mkldnn=$(func_set_params "${use_mkldnn_key}" "${use_mkldnn}")
+ set_cpu_threads=$(func_set_params "${cpu_threads_key}" "${threads}")
+ set_model_dir=$(func_set_params "${infer_model_key}" "${_model_dir}")
+ set_infer_params0=$(func_set_params "${save_log_key}" "${save_log_value}")
+ set_infer_params1=$(func_set_params "${infer_key1}" "${infer_value1}")
+ command="${_python} ${_script} ${use_gpu_key}=${use_gpu} ${set_mkldnn} ${set_cpu_threads} ${set_model_dir} ${set_batchsize} ${set_infer_params0} ${set_infer_data} ${set_benchmark} ${set_precision} ${set_infer_params1} > ${_save_log_path} 2>&1 "
+ eval $command
+ last_status=${PIPESTATUS[0]}
+ eval "cat ${_save_log_path}"
+ status_check $last_status "${command}" "${status_log}" "${model_name}"
+ done
+ done
+ done
+ done
+ elif [ ${use_gpu} = "True" ] || [ ${use_gpu} = "gpu" ]; then
+ for use_trt in ${use_trt_list[*]}; do
+ for precision in ${precision_list[*]}; do
+ if [ ${_flag_quant} = "True" ] && [ ${precision} != "int8" ]; then
+ continue
+ fi # skip when quant model inference but precision is not int8
+ for batch_size in ${batch_size_list[*]}; do
+ _save_log_path="${_log_path}/python_infer_gpu_usetrt_${use_trt}_precision_${precision}_batchsize_${batch_size}.log"
+ set_infer_data=$(func_set_params "${image_dir_key}" "${_img_dir}")
+ set_benchmark=$(func_set_params "${benchmark_key}" "${benchmark_value}")
+ set_batchsize=$(func_set_params "${batch_size_key}" "${batch_size}")
+ set_tensorrt=$(func_set_params "${use_trt_key}" "${use_trt}")
+ set_precision=$(func_set_params "${precision_key}" "${precision}")
+ set_model_dir=$(func_set_params "${infer_model_key}" "${_model_dir}")
+ set_infer_params0=$(func_set_params "${save_log_key}" "${save_log_value}")
+ set_infer_params1=$(func_set_params "${infer_key1}" "${infer_value1}")
+ command="${_python} ${_script} ${use_gpu_key}=${use_gpu} ${set_tensorrt} ${set_precision} ${set_model_dir} ${set_batchsize} ${set_infer_data} ${set_benchmark} ${set_infer_params1} ${set_infer_params0} > ${_save_log_path} 2>&1 "
+ eval $command
+ last_status=${PIPESTATUS[0]}
+ eval "cat ${_save_log_path}"
+ status_check $last_status "${command}" "${status_log}" "${model_name}"
+
+ done
+ done
+ done
+ else
+ echo "Does not support hardware other than CPU and GPU Currently!"
+ fi
+ done
+}
+
+if [ ${MODE} = "whole_infer" ]; then
+ GPUID=$3
+ if [ ${#GPUID} -le 0 ];then
+ env=" "
+ else
+ env="export CUDA_VISIBLE_DEVICES=${GPUID}"
+ fi
+ # set CUDA_VISIBLE_DEVICES
+ eval $env
+ export Count=0
+ IFS="|"
+ infer_run_exports=(${infer_export_list})
+ infer_quant_flag=(${infer_is_quant})
+ for infer_model in ${infer_model_dir_list[*]}; do
+ # run export
+ if [ ${infer_run_exports[Count]} != "null" ];then
+ save_infer_dir="${infer_model}_klquant"
+ set_export_weight=$(func_set_params "${export_weight}" "${infer_model}")
+ set_save_infer_key=$(func_set_params "${save_infer_key}" "${save_infer_dir}")
+ export_log_path="${LOG_PATH}/_export_${Count}.log"
+ export_cmd="${python} ${infer_run_exports[Count]} ${set_export_weight} ${set_save_infer_key} > ${export_log_path} 2>&1 "
+ echo ${infer_run_exports[Count]}
+ echo $export_cmd
+ eval $export_cmd
+ status_export=$?
+ status_check $status_export "${export_cmd}" "${status_log}" "${model_name}"
+ else
+ save_infer_dir=${infer_model}
+ fi
+ #run inference
+ is_quant="True"
+ func_inference "${python}" "${inference_py}" "${save_infer_dir}" "${LOG_PATH}" "${infer_img_dir}" ${is_quant}
+ Count=$(($Count + 1))
+ done
+fi
+
diff --git a/test_tipc/test_serving.sh b/test_tipc/test_serving.sh
deleted file mode 100644
index 260b252f41..0000000000
--- a/test_tipc/test_serving.sh
+++ /dev/null
@@ -1,175 +0,0 @@
-#!/bin/bash
-source test_tipc/common_func.sh
-
-FILENAME=$1
-dataline=$(awk 'NR==1, NR==18{print}' $FILENAME)
-
-# parser params
-IFS=$'\n'
-lines=(${dataline})
-
-# parser serving
-model_name=$(func_parser_value "${lines[1]}")
-python_list=$(func_parser_value "${lines[2]}")
-trans_model_py=$(func_parser_value "${lines[3]}")
-infer_model_dir_key=$(func_parser_key "${lines[4]}")
-infer_model_dir_value=$(func_parser_value "${lines[4]}")
-model_filename_key=$(func_parser_key "${lines[5]}")
-model_filename_value=$(func_parser_value "${lines[5]}")
-params_filename_key=$(func_parser_key "${lines[6]}")
-params_filename_value=$(func_parser_value "${lines[6]}")
-serving_server_key=$(func_parser_key "${lines[7]}")
-serving_server_value=$(func_parser_value "${lines[7]}")
-serving_client_key=$(func_parser_key "${lines[8]}")
-serving_client_value=$(func_parser_value "${lines[8]}")
-serving_dir_value=$(func_parser_value "${lines[9]}")
-web_service_py=$(func_parser_value "${lines[10]}")
-web_use_gpu_key=$(func_parser_key "${lines[11]}")
-web_use_gpu_list=$(func_parser_value "${lines[11]}")
-web_use_mkldnn_key=$(func_parser_key "${lines[12]}")
-web_use_mkldnn_list=$(func_parser_value "${lines[12]}")
-web_cpu_threads_key=$(func_parser_key "${lines[13]}")
-web_cpu_threads_list=$(func_parser_value "${lines[13]}")
-web_use_trt_key=$(func_parser_key "${lines[14]}")
-web_use_trt_list=$(func_parser_value "${lines[14]}")
-web_precision_key=$(func_parser_key "${lines[15]}")
-web_precision_list=$(func_parser_value "${lines[15]}")
-pipeline_py=$(func_parser_value "${lines[16]}")
-image_dir_key=$(func_parser_key "${lines[17]}")
-image_dir_value=$(func_parser_value "${lines[17]}")
-
-LOG_PATH="../../test_tipc/output"
-mkdir -p ./test_tipc/output
-status_log="${LOG_PATH}/results_serving.log"
-
-function func_serving(){
- IFS='|'
- _python=$1
- _script=$2
- _model_dir=$3
- # pdserving
- set_dirname=$(func_set_params "${infer_model_dir_key}" "${infer_model_dir_value}")
- set_model_filename=$(func_set_params "${model_filename_key}" "${model_filename_value}")
- set_params_filename=$(func_set_params "${params_filename_key}" "${params_filename_value}")
- set_serving_server=$(func_set_params "${serving_server_key}" "${serving_server_value}")
- set_serving_client=$(func_set_params "${serving_client_key}" "${serving_client_value}")
- set_image_dir=$(func_set_params "${image_dir_key}" "${image_dir_value}")
- python_list=(${python_list})
- trans_model_cmd="${python_list[0]} ${trans_model_py} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_serving_server} ${set_serving_client}"
- eval $trans_model_cmd
- cd ${serving_dir_value}
- unset https_proxy
- unset http_proxy
- for python in ${python_list[*]}; do
- if [ ${python} = "cpp" ]; then
- for use_gpu in ${web_use_gpu_list[*]}; do
- if [ ${use_gpu} = "null" ]; then
- web_service_cpp_cmd="${python_list[0]} -m paddle_serving_server.serve --model ppocr_det_mobile_2.0_serving/ ppocr_rec_mobile_2.0_serving/ --port 9293"
- eval $web_service_cpp_cmd
- last_status=${PIPESTATUS[0]}
- status_check $last_status "${web_service_cpp_cmd}" "${status_log}"
- sleep 2s
- _save_log_path="${LOG_PATH}/server_infer_cpp_cpu_pipeline_usemkldnn_False_threads_4_batchsize_1.log"
- pipeline_cmd="${python_list[0]} ocr_cpp_client.py ppocr_det_mobile_2.0_client/ ppocr_rec_mobile_2.0_client/"
- eval $pipeline_cmd
- last_status=${PIPESTATUS[0]}
- status_check $last_status "${pipeline_cmd}" "${status_log}"
- sleep 2s
- ps ux | grep -E 'web_service|pipeline' | awk '{print $2}' | xargs kill -s 9
- else
- web_service_cpp_cmd="${python_list[0]} -m paddle_serving_server.serve --model ppocr_det_mobile_2.0_serving/ ppocr_rec_mobile_2.0_serving/ --port 9293 --gpu_id=0"
- eval $web_service_cpp_cmd
- sleep 2s
- _save_log_path="${LOG_PATH}/server_infer_cpp_cpu_pipeline_usemkldnn_False_threads_4_batchsize_1.log"
- pipeline_cmd="${python_list[0]} ocr_cpp_client.py ppocr_det_mobile_2.0_client/ ppocr_rec_mobile_2.0_client/"
- eval $pipeline_cmd
- last_status=${PIPESTATUS[0]}
- status_check $last_status "${pipeline_cmd}" "${status_log}"
- sleep 2s
- ps ux | grep -E 'web_service|pipeline' | awk '{print $2}' | xargs kill -s 9
- fi
- done
- else
- # python serving
- for use_gpu in ${web_use_gpu_list[*]}; do
- if [ ${use_gpu} = "null" ]; then
- for use_mkldnn in ${web_use_mkldnn_list[*]}; do
- for threads in ${web_cpu_threads_list[*]}; do
- set_cpu_threads=$(func_set_params "${web_cpu_threads_key}" "${threads}")
- web_service_cmd="${python} ${web_service_py} ${web_use_gpu_key}="" ${web_use_mkldnn_key}=${use_mkldnn} ${set_cpu_threads} &"
- eval $web_service_cmd
- last_status=${PIPESTATUS[0]}
- status_check $last_status "${web_service_cmd}" "${status_log}"
- sleep 2s
- for pipeline in ${pipeline_py[*]}; do
- _save_log_path="${LOG_PATH}/server_infer_cpu_${pipeline%_client*}_usemkldnn_${use_mkldnn}_threads_${threads}_batchsize_1.log"
- pipeline_cmd="${python} ${pipeline} ${set_image_dir} > ${_save_log_path} 2>&1 "
- eval $pipeline_cmd
- last_status=${PIPESTATUS[0]}
- eval "cat ${_save_log_path}"
- status_check $last_status "${pipeline_cmd}" "${status_log}"
- sleep 2s
- done
- ps ux | grep -E 'web_service|pipeline' | awk '{print $2}' | xargs kill -s 9
- done
- done
- elif [ ${use_gpu} = "0" ]; then
- for use_trt in ${web_use_trt_list[*]}; do
- for precision in ${web_precision_list[*]}; do
- if [[ ${_flag_quant} = "False" ]] && [[ ${precision} =~ "int8" ]]; then
- continue
- fi
- if [[ ${precision} =~ "fp16" || ${precision} =~ "int8" ]] && [ ${use_trt} = "False" ]; then
- continue
- fi
- if [[ ${use_trt} = "False" || ${precision} =~ "int8" ]] && [[ ${_flag_quant} = "True" ]]; then
- continue
- fi
- set_tensorrt=$(func_set_params "${web_use_trt_key}" "${use_trt}")
- if [ ${use_trt} = True ]; then
- device_type=2
- fi
- set_precision=$(func_set_params "${web_precision_key}" "${precision}")
- web_service_cmd="${python} ${web_service_py} ${web_use_gpu_key}=${use_gpu} ${set_tensorrt} ${set_precision} & "
- eval $web_service_cmd
- last_status=${PIPESTATUS[0]}
- status_check $last_status "${web_service_cmd}" "${status_log}"
-
- sleep 2s
- for pipeline in ${pipeline_py[*]}; do
- _save_log_path="${LOG_PATH}/server_infer_gpu_${pipeline%_client*}_usetrt_${use_trt}_precision_${precision}_batchsize_1.log"
- pipeline_cmd="${python} ${pipeline} ${set_image_dir}> ${_save_log_path} 2>&1"
- eval $pipeline_cmd
- last_status=${PIPESTATUS[0]}
- eval "cat ${_save_log_path}"
- status_check $last_status "${pipeline_cmd}" "${status_log}"
- sleep 2s
- done
- ps ux | grep -E 'web_service|pipeline' | awk '{print $2}' | xargs kill -s 9
- done
- done
- else
- echo "Does not support hardware other than CPU and GPU Currently!"
- fi
- done
- fi
- done
-}
-
-
-#set cuda device
-GPUID=$2
-if [ ${#GPUID} -le 0 ];then
- env="export CUDA_VISIBLE_DEVICES=0"
-else
- env="export CUDA_VISIBLE_DEVICES=${GPUID}"
-fi
-eval $env
-echo $env
-
-
-echo "################### run test ###################"
-
-export Count=0
-IFS="|"
-func_serving "${web_service_cmd}"
diff --git a/test_tipc/test_serving_infer_cpp.sh b/test_tipc/test_serving_infer_cpp.sh
new file mode 100644
index 0000000000..4088c66f57
--- /dev/null
+++ b/test_tipc/test_serving_infer_cpp.sh
@@ -0,0 +1,138 @@
+#!/bin/bash
+source test_tipc/common_func.sh
+
+function func_parser_model_config(){
+ strs=$1
+ IFS="/"
+ array=(${strs})
+ tmp=${array[-1]}
+ echo ${tmp}
+}
+
+FILENAME=$1
+dataline=$(awk 'NR==1, NR==19{print}' $FILENAME)
+MODE=$2
+
+# parser params
+IFS=$'\n'
+lines=(${dataline})
+
+# parser serving
+model_name=$(func_parser_value "${lines[1]}")
+python_list=$(func_parser_value "${lines[2]}")
+trans_model_py=$(func_parser_value "${lines[3]}")
+det_infer_model_dir_key=$(func_parser_key "${lines[4]}")
+det_infer_model_dir_value=$(func_parser_value "${lines[4]}")
+model_filename_key=$(func_parser_key "${lines[5]}")
+model_filename_value=$(func_parser_value "${lines[5]}")
+params_filename_key=$(func_parser_key "${lines[6]}")
+params_filename_value=$(func_parser_value "${lines[6]}")
+det_serving_server_key=$(func_parser_key "${lines[7]}")
+det_serving_server_value=$(func_parser_value "${lines[7]}")
+det_serving_client_key=$(func_parser_key "${lines[8]}")
+det_serving_client_value=$(func_parser_value "${lines[8]}")
+rec_infer_model_dir_key=$(func_parser_key "${lines[9]}")
+rec_infer_model_dir_value=$(func_parser_value "${lines[9]}")
+rec_serving_server_key=$(func_parser_key "${lines[10]}")
+rec_serving_server_value=$(func_parser_value "${lines[10]}")
+rec_serving_client_key=$(func_parser_key "${lines[11]}")
+rec_serving_client_value=$(func_parser_value "${lines[11]}")
+det_server_value=$(func_parser_model_config "${lines[7]}")
+det_client_value=$(func_parser_model_config "${lines[8]}")
+rec_server_value=$(func_parser_model_config "${lines[10]}")
+rec_client_value=$(func_parser_model_config "${lines[11]}")
+serving_dir_value=$(func_parser_value "${lines[12]}")
+web_service_py=$(func_parser_value "${lines[13]}")
+op_key=$(func_parser_key "${lines[14]}")
+op_value=$(func_parser_value "${lines[14]}")
+port_key=$(func_parser_key "${lines[15]}")
+port_value=$(func_parser_value "${lines[15]}")
+gpu_key=$(func_parser_key "${lines[16]}")
+gpu_value=$(func_parser_value "${lines[16]}")
+cpp_client_py=$(func_parser_value "${lines[17]}")
+image_dir_key=$(func_parser_key "${lines[18]}")
+image_dir_value=$(func_parser_value "${lines[18]}")
+
+LOG_PATH="$(pwd)/test_tipc/output/${model_name}/${MODE}/cpp"
+mkdir -p ${LOG_PATH}
+status_log="${LOG_PATH}/results_cpp_serving.log"
+
+function func_serving(){
+ IFS='|'
+ _python=$1
+ _script=$2
+ _model_dir=$3
+ # pdserving
+ set_model_filename=$(func_set_params "${model_filename_key}" "${model_filename_value}")
+ set_params_filename=$(func_set_params "${params_filename_key}" "${params_filename_value}")
+ # trans det
+ set_dirname=$(func_set_params "--dirname" "${det_infer_model_dir_value}")
+ set_serving_server=$(func_set_params "--serving_server" "${det_serving_server_value}")
+ set_serving_client=$(func_set_params "--serving_client" "${det_serving_client_value}")
+ python_list=(${python_list})
+ trans_det_log="${LOG_PATH}/cpp_trans_model_det.log"
+ trans_model_cmd="${python_list[0]} ${trans_model_py} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_serving_server} ${set_serving_client} > ${trans_det_log} 2>&1 "
+ eval $trans_model_cmd
+ cp "deploy/pdserving/serving_client_conf.prototxt" ${det_serving_client_value}
+ # trans rec
+ set_dirname=$(func_set_params "--dirname" "${rec_infer_model_dir_value}")
+ set_serving_server=$(func_set_params "--serving_server" "${rec_serving_server_value}")
+ set_serving_client=$(func_set_params "--serving_client" "${rec_serving_client_value}")
+ python_list=(${python_list})
+ trans_rec_log="${LOG_PATH}/cpp_trans_model_rec.log"
+ trans_model_cmd="${python_list[0]} ${trans_model_py} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_serving_server} ${set_serving_client} > ${trans_rec_log} 2>&1 "
+ eval $trans_model_cmd
+ last_status=${PIPESTATUS[0]}
+ status_check $last_status "${trans_model_cmd}" "${status_log}" "${model_name}"
+ set_image_dir=$(func_set_params "${image_dir_key}" "${image_dir_value}")
+ python_list=(${python_list})
+ cd ${serving_dir_value}
+ # cpp serving
+ for gpu_id in ${gpu_value[*]}; do
+ if [ ${gpu_id} = "null" ]; then
+ server_log_path="${LOG_PATH}/cpp_server_cpu.log"
+ web_service_cpp_cmd="${python_list[0]} ${web_service_py} --model ${det_server_value} ${rec_server_value} ${op_key} ${op_value} ${port_key} ${port_value} > ${server_log_path} 2>&1 "
+ eval $web_service_cpp_cmd
+ last_status=${PIPESTATUS[0]}
+ status_check $last_status "${web_service_cpp_cmd}" "${status_log}" "${model_name}"
+ sleep 5s
+ _save_log_path="${LOG_PATH}/cpp_client_cpu.log"
+ cpp_client_cmd="${python_list[0]} ${cpp_client_py} ${det_client_value} ${rec_client_value} > ${_save_log_path} 2>&1"
+ eval $cpp_client_cmd
+ last_status=${PIPESTATUS[0]}
+ eval "cat ${_save_log_path}"
+ status_check $last_status "${cpp_client_cmd}" "${status_log}" "${model_name}"
+ ps ux | grep -i ${port_value} | awk '{print $2}' | xargs kill -s 9
+ else
+ server_log_path="${LOG_PATH}/cpp_server_gpu.log"
+ web_service_cpp_cmd="${python_list[0]} ${web_service_py} --model ${det_server_value} ${rec_server_value} ${op_key} ${op_value} ${port_key} ${port_value} ${gpu_key} ${gpu_id} > ${server_log_path} 2>&1 "
+ eval $web_service_cpp_cmd
+ sleep 5s
+ _save_log_path="${LOG_PATH}/cpp_client_gpu.log"
+ cpp_client_cmd="${python_list[0]} ${cpp_client_py} ${det_client_value} ${rec_client_value} > ${_save_log_path} 2>&1"
+ eval $cpp_client_cmd
+ last_status=${PIPESTATUS[0]}
+ eval "cat ${_save_log_path}"
+ status_check $last_status "${cpp_client_cmd}" "${status_log}" "${model_name}"
+ ps ux | grep -i ${port_value} | awk '{print $2}' | xargs kill -s 9
+ fi
+ done
+}
+
+
+#set cuda device
+GPUID=$3
+if [ ${#GPUID} -le 0 ];then
+ env="export CUDA_VISIBLE_DEVICES=0"
+else
+ env="export CUDA_VISIBLE_DEVICES=${GPUID}"
+fi
+eval $env
+echo $env
+
+
+echo "################### run test ###################"
+
+export Count=0
+IFS="|"
+func_serving "${web_service_cpp_cmd}"
diff --git a/test_tipc/test_serving_infer_python.sh b/test_tipc/test_serving_infer_python.sh
new file mode 100644
index 0000000000..57dab6aeb5
--- /dev/null
+++ b/test_tipc/test_serving_infer_python.sh
@@ -0,0 +1,229 @@
+#!/bin/bash
+source test_tipc/common_func.sh
+
+function func_parser_model_config(){
+ strs=$1
+ IFS="/"
+ array=(${strs})
+ tmp=${array[-1]}
+ echo ${tmp}
+}
+
+FILENAME=$1
+dataline=$(awk 'NR==1, NR==23{print}' $FILENAME)
+MODE=$2
+
+# parser params
+IFS=$'\n'
+lines=(${dataline})
+
+# parser serving
+model_name=$(func_parser_value "${lines[1]}")
+python_list=$(func_parser_value "${lines[2]}")
+trans_model_py=$(func_parser_value "${lines[3]}")
+det_infer_model_dir_key=$(func_parser_key "${lines[4]}")
+det_infer_model_dir_value=$(func_parser_value "${lines[4]}")
+model_filename_key=$(func_parser_key "${lines[5]}")
+model_filename_value=$(func_parser_value "${lines[5]}")
+params_filename_key=$(func_parser_key "${lines[6]}")
+params_filename_value=$(func_parser_value "${lines[6]}")
+det_serving_server_key=$(func_parser_key "${lines[7]}")
+det_serving_server_value=$(func_parser_value "${lines[7]}")
+det_serving_client_key=$(func_parser_key "${lines[8]}")
+det_serving_client_value=$(func_parser_value "${lines[8]}")
+rec_infer_model_dir_key=$(func_parser_key "${lines[9]}")
+rec_infer_model_dir_value=$(func_parser_value "${lines[9]}")
+rec_serving_server_key=$(func_parser_key "${lines[10]}")
+rec_serving_server_value=$(func_parser_value "${lines[10]}")
+rec_serving_client_key=$(func_parser_key "${lines[11]}")
+rec_serving_client_value=$(func_parser_value "${lines[11]}")
+serving_dir_value=$(func_parser_value "${lines[12]}")
+web_service_py=$(func_parser_value "${lines[13]}")
+web_use_gpu_key=$(func_parser_key "${lines[14]}")
+web_use_gpu_list=$(func_parser_value "${lines[14]}")
+web_use_mkldnn_key=$(func_parser_key "${lines[15]}")
+web_use_mkldnn_list=$(func_parser_value "${lines[15]}")
+web_cpu_threads_key=$(func_parser_key "${lines[16]}")
+web_cpu_threads_list=$(func_parser_value "${lines[16]}")
+web_use_trt_key=$(func_parser_key "${lines[17]}")
+web_use_trt_list=$(func_parser_value "${lines[17]}")
+web_precision_key=$(func_parser_key "${lines[18]}")
+web_precision_list=$(func_parser_value "${lines[18]}")
+det_server_key=$(func_parser_key "${lines[19]}")
+det_server_value=$(func_parser_model_config "${lines[7]}")
+det_client_value=$(func_parser_model_config "${lines[8]}")
+rec_server_key=$(func_parser_key "${lines[20]}")
+rec_server_value=$(func_parser_model_config "${lines[10]}")
+rec_client_value=$(func_parser_model_config "${lines[11]}")
+pipeline_py=$(func_parser_value "${lines[21]}")
+image_dir_key=$(func_parser_key "${lines[22]}")
+image_dir_value=$(func_parser_value "${lines[22]}")
+
+LOG_PATH="$(pwd)/test_tipc/output/${model_name}/${MODE}/python"
+mkdir -p ${LOG_PATH}
+status_log="${LOG_PATH}/results_python_serving.log"
+
+function func_serving(){
+ IFS='|'
+ _python=$1
+ _script=$2
+ _model_dir=$3
+ # pdserving
+ set_model_filename=$(func_set_params "${model_filename_key}" "${model_filename_value}")
+ set_params_filename=$(func_set_params "${params_filename_key}" "${params_filename_value}")
+ if [ ${model_name} = "ch_PP-OCRv2" ] || [ ${model_name} = "ch_PP-OCRv3" ] || [ ${model_name} = "ch_ppocr_mobile_v2.0" ] || [ ${model_name} = "ch_ppocr_server_v2.0" ]; then
+ # trans det
+ set_dirname=$(func_set_params "--dirname" "${det_infer_model_dir_value}")
+ set_serving_server=$(func_set_params "--serving_server" "${det_serving_server_value}")
+ set_serving_client=$(func_set_params "--serving_client" "${det_serving_client_value}")
+ python_list=(${python_list})
+ trans_det_log="${LOG_PATH}/python_trans_model_det.log"
+ trans_model_cmd="${python_list[0]} ${trans_model_py} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_serving_server} ${set_serving_client} > ${trans_det_log} 2>&1 "
+ eval $trans_model_cmd
+ # trans rec
+ set_dirname=$(func_set_params "--dirname" "${rec_infer_model_dir_value}")
+ set_serving_server=$(func_set_params "--serving_server" "${rec_serving_server_value}")
+ set_serving_client=$(func_set_params "--serving_client" "${rec_serving_client_value}")
+ python_list=(${python_list})
+ trans_rec_log="${LOG_PATH}/python_trans_model_rec.log"
+ trans_model_cmd="${python_list[0]} ${trans_model_py} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_serving_server} ${set_serving_client} > ${trans_rec_log} 2>&1 "
+ eval $trans_model_cmd
+ elif [[ ${model_name} =~ "det" ]]; then
+ # trans det
+ set_dirname=$(func_set_params "--dirname" "${det_infer_model_dir_value}")
+ set_serving_server=$(func_set_params "--serving_server" "${det_serving_server_value}")
+ set_serving_client=$(func_set_params "--serving_client" "${det_serving_client_value}")
+ python_list=(${python_list})
+ trans_det_log="${LOG_PATH}/python_trans_model_det.log"
+ trans_model_cmd="${python_list[0]} ${trans_model_py} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_serving_server} ${set_serving_client} > ${trans_det_log} 2>&1 "
+ eval $trans_model_cmd
+ elif [[ ${model_name} =~ "rec" ]]; then
+ # trans rec
+ set_dirname=$(func_set_params "--dirname" "${rec_infer_model_dir_value}")
+ set_serving_server=$(func_set_params "--serving_server" "${rec_serving_server_value}")
+ set_serving_client=$(func_set_params "--serving_client" "${rec_serving_client_value}")
+ python_list=(${python_list})
+ trans_rec_log="${LOG_PATH}/python_trans_model_rec.log"
+ trans_model_cmd="${python_list[0]} ${trans_model_py} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_serving_server} ${set_serving_client} > ${trans_rec_log} 2>&1 "
+ eval $trans_model_cmd
+ fi
+ set_image_dir=$(func_set_params "${image_dir_key}" "${image_dir_value}")
+ python_list=(${python_list})
+
+ cd ${serving_dir_value}
+ python=${python_list[0]}
+
+ # python serving
+ for use_gpu in ${web_use_gpu_list[*]}; do
+ if [ ${use_gpu} = "null" ]; then
+ for use_mkldnn in ${web_use_mkldnn_list[*]}; do
+ for threads in ${web_cpu_threads_list[*]}; do
+ set_cpu_threads=$(func_set_params "${web_cpu_threads_key}" "${threads}")
+ server_log_path="${LOG_PATH}/python_server_cpu_usemkldnn_${use_mkldnn}_threads_${threads}.log"
+ if [ ${model_name} = "ch_PP-OCRv2" ] || [ ${model_name} = "ch_PP-OCRv3" ] || [ ${model_name} = "ch_ppocr_mobile_v2.0" ] || [ ${model_name} = "ch_ppocr_server_v2.0" ]; then
+ set_det_model_config=$(func_set_params "${det_server_key}" "${det_server_value}")
+ set_rec_model_config=$(func_set_params "${rec_server_key}" "${rec_server_value}")
+ web_service_cmd="${python} ${web_service_py} ${web_use_gpu_key}="" ${web_use_mkldnn_key}=${use_mkldnn} ${set_cpu_threads} ${set_det_model_config} ${set_rec_model_config} > ${server_log_path} 2>&1 "
+ eval $web_service_cmd
+ last_status=${PIPESTATUS[0]}
+ status_check $last_status "${web_service_cmd}" "${status_log}" "${model_name}"
+ elif [[ ${model_name} =~ "det" ]]; then
+ set_det_model_config=$(func_set_params "${det_server_key}" "${det_server_value}")
+ web_service_cmd="${python} ${web_service_py} ${web_use_gpu_key}="" ${web_use_mkldnn_key}=${use_mkldnn} ${set_cpu_threads} ${set_det_model_config} > ${server_log_path} 2>&1 "
+ eval $web_service_cmd
+ last_status=${PIPESTATUS[0]}
+ status_check $last_status "${web_service_cmd}" "${status_log}" "${model_name}"
+ elif [[ ${model_name} =~ "rec" ]]; then
+ set_rec_model_config=$(func_set_params "${rec_server_key}" "${rec_server_value}")
+ web_service_cmd="${python} ${web_service_py} ${web_use_gpu_key}="" ${web_use_mkldnn_key}=${use_mkldnn} ${set_cpu_threads} ${set_rec_model_config} > ${server_log_path} 2>&1 "
+ eval $web_service_cmd
+ last_status=${PIPESTATUS[0]}
+ status_check $last_status "${web_service_cmd}" "${status_log}" "${model_name}"
+ fi
+ sleep 2s
+ for pipeline in ${pipeline_py[*]}; do
+ _save_log_path="${LOG_PATH}/python_client_cpu_${pipeline%_client*}_usemkldnn_${use_mkldnn}_threads_${threads}_batchsize_1.log"
+ pipeline_cmd="${python} ${pipeline} ${set_image_dir} > ${_save_log_path} 2>&1 "
+ eval $pipeline_cmd
+ last_status=${PIPESTATUS[0]}
+ eval "cat ${_save_log_path}"
+ status_check $last_status "${pipeline_cmd}" "${status_log}" "${model_name}"
+ sleep 2s
+ done
+ ps ux | grep -E 'web_service' | awk '{print $2}' | xargs kill -s 9
+ done
+ done
+ elif [ ${use_gpu} = "gpu" ]; then
+ for use_trt in ${web_use_trt_list[*]}; do
+ for precision in ${web_precision_list[*]}; do
+ server_log_path="${LOG_PATH}/python_server_gpu_usetrt_${use_trt}_precision_${precision}.log"
+ if [[ ${_flag_quant} = "False" ]] && [[ ${precision} =~ "int8" ]]; then
+ continue
+ fi
+ if [[ ${precision} =~ "fp16" || ${precision} =~ "int8" ]] && [ ${use_trt} = "False" ]; then
+ continue
+ fi
+ if [[ ${use_trt} = "False" || ${precision} =~ "int8" ]] && [[ ${_flag_quant} = "True" ]]; then
+ continue
+ fi
+ set_tensorrt=$(func_set_params "${web_use_trt_key}" "${use_trt}")
+ if [ ${use_trt} = True ]; then
+ device_type=2
+ fi
+ set_precision=$(func_set_params "${web_precision_key}" "${precision}")
+ if [ ${model_name} = "ch_PP-OCRv2" ] || [ ${model_name} = "ch_PP-OCRv3" ] || [ ${model_name} = "ch_ppocr_mobile_v2.0" ] || [ ${model_name} = "ch_ppocr_server_v2.0" ]; then
+ set_det_model_config=$(func_set_params "${det_server_key}" "${det_server_value}")
+ set_rec_model_config=$(func_set_params "${rec_server_key}" "${rec_server_value}")
+ web_service_cmd="${python} ${web_service_py} ${set_tensorrt} ${set_precision} ${set_det_model_config} ${set_rec_model_config} > ${server_log_path} 2>&1 "
+ eval $web_service_cmd
+ last_status=${PIPESTATUS[0]}
+ status_check $last_status "${web_service_cmd}" "${status_log}" "${model_name}"
+ elif [[ ${model_name} =~ "det" ]]; then
+ set_det_model_config=$(func_set_params "${det_server_key}" "${det_server_value}")
+ web_service_cmd="${python} ${web_service_py} ${set_tensorrt} ${set_precision} ${set_det_model_config} > ${server_log_path} 2>&1 "
+ eval $web_service_cmd
+ last_status=${PIPESTATUS[0]}
+ status_check $last_status "${web_service_cmd}" "${status_log}" "${model_name}"
+ elif [[ ${model_name} =~ "rec" ]]; then
+ set_rec_model_config=$(func_set_params "${rec_server_key}" "${rec_server_value}")
+ web_service_cmd="${python} ${web_service_py} ${set_tensorrt} ${set_precision} ${set_rec_model_config} > ${server_log_path} 2>&1 "
+ eval $web_service_cmd
+ last_status=${PIPESTATUS[0]}
+ status_check $last_status "${web_service_cmd}" "${status_log}" "${model_name}"
+ fi
+ sleep 2s
+ for pipeline in ${pipeline_py[*]}; do
+ _save_log_path="${LOG_PATH}/python_client_gpu_${pipeline%_client*}_usetrt_${use_trt}_precision_${precision}_batchsize_1.log"
+ pipeline_cmd="${python} ${pipeline} ${set_image_dir}> ${_save_log_path} 2>&1"
+ eval $pipeline_cmd
+ last_status=${PIPESTATUS[0]}
+ eval "cat ${_save_log_path}"
+ status_check $last_status "${pipeline_cmd}" "${status_log}" "${model_name}"
+ sleep 2s
+ done
+ ps ux | grep -E 'web_service' | awk '{print $2}' | xargs kill -s 9
+ done
+ done
+ else
+ echo "Does not support hardware other than CPU and GPU Currently!"
+ fi
+ done
+}
+
+
+#set cuda device
+GPUID=$3
+if [ ${#GPUID} -le 0 ];then
+ env="export CUDA_VISIBLE_DEVICES=0"
+else
+ env="export CUDA_VISIBLE_DEVICES=${GPUID}"
+fi
+eval $env
+echo $env
+
+
+echo "################### run test ###################"
+
+export Count=0
+IFS="|"
+func_serving "${web_service_cmd}"
diff --git a/test_tipc/test_train_inference_python.sh b/test_tipc/test_train_inference_python.sh
index fe98cb00f6..fa68cb2632 100644
--- a/test_tipc/test_train_inference_python.sh
+++ b/test_tipc/test_train_inference_python.sh
@@ -2,7 +2,7 @@
source test_tipc/common_func.sh
FILENAME=$1
-# MODE be one of ['lite_train_lite_infer' 'lite_train_whole_infer' 'whole_train_whole_infer', 'whole_infer', 'klquant_whole_infer']
+# MODE be one of ['lite_train_lite_infer' 'lite_train_whole_infer' 'whole_train_whole_infer', 'whole_infer']
MODE=$2
dataline=$(awk 'NR==1, NR==51{print}' $FILENAME)
@@ -88,44 +88,7 @@ benchmark_value=$(func_parser_value "${lines[49]}")
infer_key1=$(func_parser_key "${lines[50]}")
infer_value1=$(func_parser_value "${lines[50]}")
-# parser klquant_infer
-if [ ${MODE} = "klquant_whole_infer" ]; then
- dataline=$(awk 'NR==1, NR==17{print}' $FILENAME)
- lines=(${dataline})
- model_name=$(func_parser_value "${lines[1]}")
- python=$(func_parser_value "${lines[2]}")
- export_weight=$(func_parser_key "${lines[3]}")
- save_infer_key=$(func_parser_key "${lines[4]}")
- # parser inference model
- infer_model_dir_list=$(func_parser_value "${lines[5]}")
- infer_export_list=$(func_parser_value "${lines[6]}")
- infer_is_quant=$(func_parser_value "${lines[7]}")
- # parser inference
- inference_py=$(func_parser_value "${lines[8]}")
- use_gpu_key=$(func_parser_key "${lines[9]}")
- use_gpu_list=$(func_parser_value "${lines[9]}")
- use_mkldnn_key=$(func_parser_key "${lines[10]}")
- use_mkldnn_list=$(func_parser_value "${lines[10]}")
- cpu_threads_key=$(func_parser_key "${lines[11]}")
- cpu_threads_list=$(func_parser_value "${lines[11]}")
- batch_size_key=$(func_parser_key "${lines[12]}")
- batch_size_list=$(func_parser_value "${lines[12]}")
- use_trt_key=$(func_parser_key "${lines[13]}")
- use_trt_list=$(func_parser_value "${lines[13]}")
- precision_key=$(func_parser_key "${lines[14]}")
- precision_list=$(func_parser_value "${lines[14]}")
- infer_model_key=$(func_parser_key "${lines[15]}")
- image_dir_key=$(func_parser_key "${lines[16]}")
- infer_img_dir=$(func_parser_value "${lines[16]}")
- save_log_key=$(func_parser_key "${lines[17]}")
- save_log_value=$(func_parser_value "${lines[17]}")
- benchmark_key=$(func_parser_key "${lines[18]}")
- benchmark_value=$(func_parser_value "${lines[18]}")
- infer_key1=$(func_parser_key "${lines[19]}")
- infer_value1=$(func_parser_value "${lines[19]}")
-fi
-
-LOG_PATH="./test_tipc/output/${model_name}"
+LOG_PATH="./test_tipc/output/${model_name}/${MODE}"
mkdir -p ${LOG_PATH}
status_log="${LOG_PATH}/results_python.log"
@@ -142,9 +105,9 @@ function func_inference(){
for use_gpu in ${use_gpu_list[*]}; do
if [ ${use_gpu} = "False" ] || [ ${use_gpu} = "cpu" ]; then
for use_mkldnn in ${use_mkldnn_list[*]}; do
- if [ ${use_mkldnn} = "False" ] && [ ${_flag_quant} = "True" ]; then
- continue
- fi
+ # if [ ${use_mkldnn} = "False" ] && [ ${_flag_quant} = "True" ]; then
+ # continue
+ # fi
for threads in ${cpu_threads_list[*]}; do
for batch_size in ${batch_size_list[*]}; do
for precision in ${precision_list[*]}; do
@@ -169,7 +132,7 @@ function func_inference(){
eval $command
last_status=${PIPESTATUS[0]}
eval "cat ${_save_log_path}"
- status_check $last_status "${command}" "${status_log}"
+ status_check $last_status "${command}" "${status_log}" "${model_name}"
done
done
done
@@ -200,7 +163,7 @@ function func_inference(){
eval $command
last_status=${PIPESTATUS[0]}
eval "cat ${_save_log_path}"
- status_check $last_status "${command}" "${status_log}"
+ status_check $last_status "${command}" "${status_log}" "${model_name}"
done
done
@@ -211,7 +174,7 @@ function func_inference(){
done
}
-if [ ${MODE} = "whole_infer" ] || [ ${MODE} = "klquant_whole_infer" ]; then
+if [ ${MODE} = "whole_infer" ]; then
GPUID=$3
if [ ${#GPUID} -le 0 ];then
env=" "
@@ -226,29 +189,22 @@ if [ ${MODE} = "whole_infer" ] || [ ${MODE} = "klquant_whole_infer" ]; then
infer_quant_flag=(${infer_is_quant})
for infer_model in ${infer_model_dir_list[*]}; do
# run export
- if [ ${infer_run_exports[Count]} != "null" ];then
- if [ ${MODE} = "klquant_whole_infer" ]; then
- save_infer_dir="${infer_model}_klquant"
- fi
- if [ ${MODE} = "whole_infer" ]; then
- save_infer_dir="${infer_model}"
- fi
+ if [ ${infer_run_exports[Count]} != "null" ];then
+ save_infer_dir="${infer_model}"
set_export_weight=$(func_set_params "${export_weight}" "${infer_model}")
set_save_infer_key=$(func_set_params "${save_infer_key}" "${save_infer_dir}")
- export_cmd="${python} ${infer_run_exports[Count]} ${set_export_weight} ${set_save_infer_key}"
+ export_log_path="${LOG_PATH}/_export_${Count}.log"
+ export_cmd="${python} ${infer_run_exports[Count]} ${set_export_weight} ${set_save_infer_key} > ${export_log_path} 2>&1 "
echo ${infer_run_exports[Count]}
echo $export_cmd
eval $export_cmd
status_export=$?
- status_check $status_export "${export_cmd}" "${status_log}"
+ status_check $status_export "${export_cmd}" "${status_log}" "${model_name}"
else
save_infer_dir=${infer_model}
fi
#run inference
is_quant=${infer_quant_flag[Count]}
- if [ ${MODE} = "klquant_whole_infer" ]; then
- is_quant="True"
- fi
func_inference "${python}" "${inference_py}" "${save_infer_dir}" "${LOG_PATH}" "${infer_img_dir}" ${is_quant}
Count=$(($Count + 1))
done
@@ -315,7 +271,9 @@ else
set_batchsize=$(func_set_params "${train_batch_key}" "${train_batch_value}")
set_train_params1=$(func_set_params "${train_param_key1}" "${train_param_value1}")
set_use_gpu=$(func_set_params "${train_use_gpu_key}" "${train_use_gpu}")
- if [ ${#ips} -le 26 ];then
+ # if length of ips >= 15, then it is seen as multi-machine
+ # 15 is the min length of ips info for multi-machine: 0.0.0.0,0.0.0.0
+ if [ ${#ips} -le 15 ];then
save_log="${LOG_PATH}/${trainer}_gpus_${gpu}_autocast_${autocast}"
nodes=1
else
@@ -330,14 +288,14 @@ else
set_save_model=$(func_set_params "${save_model_key}" "${save_log}")
if [ ${#gpu} -le 2 ];then # train with cpu or single gpu
cmd="${python} ${run_train} ${set_use_gpu} ${set_save_model} ${set_epoch} ${set_pretrain} ${set_autocast} ${set_batchsize} ${set_train_params1} ${set_amp_config} "
- elif [ ${#ips} -le 26 ];then # train with multi-gpu
+ elif [ ${#ips} -le 15 ];then # train with multi-gpu
cmd="${python} -m paddle.distributed.launch --gpus=${gpu} ${run_train} ${set_use_gpu} ${set_save_model} ${set_epoch} ${set_pretrain} ${set_autocast} ${set_batchsize} ${set_train_params1} ${set_amp_config}"
else # train with multi-machine
cmd="${python} -m paddle.distributed.launch --ips=${ips} --gpus=${gpu} ${run_train} ${set_use_gpu} ${set_save_model} ${set_pretrain} ${set_epoch} ${set_autocast} ${set_batchsize} ${set_train_params1} ${set_amp_config}"
fi
# run train
eval $cmd
- status_check $? "${cmd}" "${status_log}"
+ status_check $? "${cmd}" "${status_log}" "${model_name}"
set_eval_pretrain=$(func_set_params "${pretrain_model_key}" "${save_log}/${train_model_name}")
@@ -345,19 +303,21 @@ else
if [ ${eval_py} != "null" ]; then
eval ${env}
set_eval_params1=$(func_set_params "${eval_key1}" "${eval_value1}")
- eval_cmd="${python} ${eval_py} ${set_eval_pretrain} ${set_use_gpu} ${set_eval_params1}"
+ eval_log_path="${LOG_PATH}/${trainer}_gpus_${gpu}_autocast_${autocast}_nodes_${nodes}_eval.log"
+ eval_cmd="${python} ${eval_py} ${set_eval_pretrain} ${set_use_gpu} ${set_eval_params1} > ${eval_log_path} 2>&1 "
eval $eval_cmd
- status_check $? "${eval_cmd}" "${status_log}"
+ status_check $? "${eval_cmd}" "${status_log}" "${model_name}"
fi
# run export model
if [ ${run_export} != "null" ]; then
# run export model
save_infer_path="${save_log}"
+ export_log_path="${LOG_PATH}/${trainer}_gpus_${gpu}_autocast_${autocast}_nodes_${nodes}_export.log"
set_export_weight=$(func_set_params "${export_weight}" "${save_log}/${train_model_name}")
set_save_infer_key=$(func_set_params "${save_infer_key}" "${save_infer_path}")
- export_cmd="${python} ${run_export} ${set_export_weight} ${set_save_infer_key}"
+ export_cmd="${python} ${run_export} ${set_export_weight} ${set_save_infer_key} > ${export_log_path} 2>&1 "
eval $export_cmd
- status_check $? "${export_cmd}" "${status_log}"
+ status_check $? "${export_cmd}" "${status_log}" "${model_name}"
#run inference
eval $env
diff --git a/tools/end2end/convert_ppocr_label.py b/tools/end2end/convert_ppocr_label.py
index 8084cac785..c64b9ed168 100644
--- a/tools/end2end/convert_ppocr_label.py
+++ b/tools/end2end/convert_ppocr_label.py
@@ -85,10 +85,16 @@ def convert_label(label_dir, mode="gt", save_dir="./save_results/"):
print("The convert label saved in {}".format(save_dir))
+def parse_args():
+ import argparse
+ parser = argparse.ArgumentParser(description="args")
+ parser.add_argument("--label_path", type=str, required=True)
+ parser.add_argument("--save_folder", type=str, required=True)
+ parser.add_argument("--mode", type=str, default=False)
+ args = parser.parse_args()
+ return args
+
+
if __name__ == "__main__":
-
- ppocr_label_gt = "/paddle/Datasets/chinese/test_set/Label_refine_310_V2.txt"
- convert_label(ppocr_label_gt, "gt", "./save_gt_310_V2/")
-
- ppocr_label_gt = "./infer_results/ch_PPOCRV2_infer.txt"
- convert_label(ppocr_label_gt_en, "pred", "./save_PPOCRV2_infer/")
+ args = parse_args()
+ convert_label(args.label_path, args.mode, args.save_folder)
diff --git a/tools/end2end/readme.md b/tools/end2end/readme.md
index 69da06dcda..636ee764ae 100644
--- a/tools/end2end/readme.md
+++ b/tools/end2end/readme.md
@@ -23,19 +23,13 @@ all-sum-510/00224225.jpg [{"transcription": "超赞", "points": [[8.0, 48
**步骤二：**
将步骤一保存的数据转换为端对端评测需要的数据格式：
-修改 `tools/convert_ppocr_label.py`中的代码，convert_label函数中设置输入标签路径，Mode，保存标签路径等，对预测数据的GTlabel和预测结果的label格式进行转换。
+
+修改 `tools/end2end/convert_ppocr_label.py`中的代码，convert_label函数中设置输入标签路径，Mode，保存标签路径等，对预测数据的GTlabel和预测结果的label格式进行转换。
```
-ppocr_label_gt = "gt_label.txt"
-convert_label(ppocr_label_gt, "gt", "./save_gt_label/")
+python3 tools/end2end/convert_ppocr_label.py --mode=gt --label_path=path/to/label_txt --save_folder=save_gt_label
-ppocr_label_gt = "./ch_PP-OCRv2_results/system_results.txt"
-convert_label(ppocr_label_gt_en, "pred", "./save_PPOCRV2_infer/")
-```
-
-运行`convert_ppocr_label.py`:
-```
-python3 tools/convert_ppocr_label.py
+python3 tools/end2end/convert_ppocr_label.py --mode=pred --label_path=path/to/pred_txt --save_folder=save_PPOCRV2_infer
```
得到如下结果：
diff --git a/tools/export_model.py b/tools/export_model.py
index e971f6cb20..b10d41d5b2 100755
--- a/tools/export_model.py
+++ b/tools/export_model.py
@@ -17,7 +17,7 @@ import sys
__dir__ = os.path.dirname(os.path.abspath(__file__))
sys.path.append(__dir__)
-sys.path.append(os.path.abspath(os.path.join(__dir__, "..")))
+sys.path.insert(0, os.path.abspath(os.path.join(__dir__, "..")))
import argparse
@@ -31,7 +31,12 @@ from ppocr.utils.logging import get_logger
from tools.program import load_config, merge_config, ArgsParser
-def export_single_model(model, arch_config, save_path, logger, quanter=None):
+def export_single_model(model,
+ arch_config,
+ save_path,
+ logger,
+ input_shape=None,
+ quanter=None):
if arch_config["algorithm"] == "SRN":
max_text_length = arch_config["Head"]["max_text_length"]
other_shape = [
@@ -64,7 +69,7 @@ def export_single_model(model, arch_config, save_path, logger, quanter=None):
else:
other_shape = [
paddle.static.InputSpec(
- shape=[None, 3, 64, 256], dtype="float32"),
+ shape=[None] + input_shape, dtype="float32"),
]
model = to_static(model, input_spec=other_shape)
elif arch_config["algorithm"] == "PREN":
@@ -73,10 +78,29 @@ def export_single_model(model, arch_config, save_path, logger, quanter=None):
shape=[None, 3, 64, 512], dtype="float32"),
]
model = to_static(model, input_spec=other_shape)
+ elif arch_config["algorithm"] == "ViTSTR":
+ other_shape = [
+ paddle.static.InputSpec(
+ shape=[None, 1, 224, 224], dtype="float32"),
+ ]
+ model = to_static(model, input_spec=other_shape)
+ elif arch_config["algorithm"] == "ABINet":
+ other_shape = [
+ paddle.static.InputSpec(
+ shape=[None, 3, 32, 128], dtype="float32"),
+ ]
+ # print([None, 3, 32, 128])
+ model = to_static(model, input_spec=other_shape)
+ elif arch_config["algorithm"] == "NRTR":
+ other_shape = [
+ paddle.static.InputSpec(
+ shape=[None, 1, 32, 100], dtype="float32"),
+ ]
+ model = to_static(model, input_spec=other_shape)
else:
infer_shape = [3, -1, -1]
if arch_config["model_type"] == "rec":
- infer_shape = [3, 32, -1] # for rec model, H must be 32
+ infer_shape = [3, 48, -1] # for rec model, H must be 32
if "Transform" in arch_config and arch_config[
"Transform"] is not None and arch_config["Transform"][
"name"] == "TPS":
@@ -84,8 +108,6 @@ def export_single_model(model, arch_config, save_path, logger, quanter=None):
"When there is tps in the network, variable length input is not supported, and the input size needs to be the same as during training"
)
infer_shape[-1] = 100
- if arch_config["algorithm"] == "NRTR":
- infer_shape = [1, 32, 100]
elif arch_config["model_type"] == "table":
infer_shape = [3, 488, 488]
model = to_static(
@@ -157,6 +179,13 @@ def main():
arch_config = config["Architecture"]
+ if arch_config["algorithm"] == "SVTR" and arch_config["Head"][
+ "name"] != 'MultiHead':
+ input_shape = config["Eval"]["dataset"]["transforms"][-2][
+ 'SVTRRecResizeImg']['image_shape']
+ else:
+ input_shape = None
+
if arch_config["algorithm"] in ["Distillation", ]: # distillation model
archs = list(arch_config["Models"].values())
for idx, name in enumerate(model.model_name_list):
@@ -165,7 +194,8 @@ def main():
sub_model_save_path, logger)
else:
save_path = os.path.join(save_path, "inference")
- export_single_model(model, arch_config, save_path, logger)
+ export_single_model(
+ model, arch_config, save_path, logger, input_shape=input_shape)
if __name__ == "__main__":
diff --git a/tools/infer/predict_det.py b/tools/infer/predict_det.py
index 5f2675d667..7b6bebf1fb 100755
--- a/tools/infer/predict_det.py
+++ b/tools/infer/predict_det.py
@@ -154,9 +154,10 @@ class TextDetector(object):
s = pts.sum(axis=1)
rect[0] = pts[np.argmin(s)]
rect[2] = pts[np.argmax(s)]
- diff = np.diff(pts, axis=1)
- rect[1] = pts[np.argmin(diff)]
- rect[3] = pts[np.argmax(diff)]
+ tmp = np.delete(pts, (np.argmin(s), np.argmax(s)), axis=0)
+ diff = np.diff(np.array(tmp), axis=1)
+ rect[1] = tmp[np.argmin(diff)]
+ rect[3] = tmp[np.argmax(diff)]
return rect
def clip_det_res(self, points, img_height, img_width):
diff --git a/tools/infer/predict_rec.py b/tools/infer/predict_rec.py
index 3664ef2caf..a95f555966 100755
--- a/tools/infer/predict_rec.py
+++ b/tools/infer/predict_rec.py
@@ -69,6 +69,18 @@ class TextRecognizer(object):
"character_dict_path": args.rec_char_dict_path,
"use_space_char": args.use_space_char
}
+ elif self.rec_algorithm == 'ViTSTR':
+ postprocess_params = {
+ 'name': 'ViTSTRLabelDecode',
+ "character_dict_path": args.rec_char_dict_path,
+ "use_space_char": args.use_space_char
+ }
+ elif self.rec_algorithm == 'ABINet':
+ postprocess_params = {
+ 'name': 'ABINetLabelDecode',
+ "character_dict_path": args.rec_char_dict_path,
+ "use_space_char": args.use_space_char
+ }
self.postprocess_op = build_post_process(postprocess_params)
self.predictor, self.input_tensor, self.output_tensors, self.config = \
utility.create_predictor(args, 'rec', logger)
@@ -96,15 +108,22 @@ class TextRecognizer(object):
def resize_norm_img(self, img, max_wh_ratio):
imgC, imgH, imgW = self.rec_image_shape
- if self.rec_algorithm == 'NRTR':
+ if self.rec_algorithm == 'NRTR' or self.rec_algorithm == 'ViTSTR':
img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
# return padding_im
image_pil = Image.fromarray(np.uint8(img))
- img = image_pil.resize([100, 32], Image.ANTIALIAS)
+ if self.rec_algorithm == 'ViTSTR':
+ img = image_pil.resize([imgW, imgH], Image.BICUBIC)
+ else:
+ img = image_pil.resize([imgW, imgH], Image.ANTIALIAS)
img = np.array(img)
norm_img = np.expand_dims(img, -1)
norm_img = norm_img.transpose((2, 0, 1))
- return norm_img.astype(np.float32) / 128. - 1.
+ if self.rec_algorithm == 'ViTSTR':
+ norm_img = norm_img.astype(np.float32) / 255.
+ else:
+ norm_img = norm_img.astype(np.float32) / 128. - 1.
+ return norm_img
assert imgC == img.shape[2]
imgW = int((imgH * max_wh_ratio))
@@ -132,17 +151,6 @@ class TextRecognizer(object):
padding_im[:, :, 0:resized_w] = resized_image
return padding_im
- def resize_norm_img_svtr(self, img, image_shape):
-
- imgC, imgH, imgW = image_shape
- resized_image = cv2.resize(
- img, (imgW, imgH), interpolation=cv2.INTER_LINEAR)
- resized_image = resized_image.astype('float32')
- resized_image = resized_image.transpose((2, 0, 1)) / 255
- resized_image -= 0.5
- resized_image /= 0.5
- return resized_image
-
def resize_norm_img_srn(self, img, image_shape):
imgC, imgH, imgW = image_shape
@@ -250,6 +258,35 @@ class TextRecognizer(object):
return padding_im, resize_shape, pad_shape, valid_ratio
+ def resize_norm_img_svtr(self, img, image_shape):
+
+ imgC, imgH, imgW = image_shape
+ resized_image = cv2.resize(
+ img, (imgW, imgH), interpolation=cv2.INTER_LINEAR)
+ resized_image = resized_image.astype('float32')
+ resized_image = resized_image.transpose((2, 0, 1)) / 255
+ resized_image -= 0.5
+ resized_image /= 0.5
+ return resized_image
+
+ def resize_norm_img_abinet(self, img, image_shape):
+
+ imgC, imgH, imgW = image_shape
+
+ resized_image = cv2.resize(
+ img, (imgW, imgH), interpolation=cv2.INTER_LINEAR)
+ resized_image = resized_image.astype('float32')
+ resized_image = resized_image / 255.
+
+ mean = np.array([0.485, 0.456, 0.406])
+ std = np.array([0.229, 0.224, 0.225])
+ resized_image = (
+ resized_image - mean[None, None, ...]) / std[None, None, ...]
+ resized_image = resized_image.transpose((2, 0, 1))
+ resized_image = resized_image.astype('float32')
+
+ return resized_image
+
def __call__(self, img_list):
img_num = len(img_list)
# Calculate the aspect ratio of all text bars
@@ -300,6 +337,11 @@ class TextRecognizer(object):
self.rec_image_shape)
norm_img = norm_img[np.newaxis, :]
norm_img_batch.append(norm_img)
+ elif self.rec_algorithm == "ABINet":
+ norm_img = self.resize_norm_img_abinet(
+ img_list[indices[ino]], self.rec_image_shape)
+ norm_img = norm_img[np.newaxis, :]
+ norm_img_batch.append(norm_img)
else:
norm_img = self.resize_norm_img(img_list[indices[ino]],
max_wh_ratio)
diff --git a/tools/infer/utility.py b/tools/infer/utility.py
index 74ec42ec84..366212f228 100644
--- a/tools/infer/utility.py
+++ b/tools/infer/utility.py
@@ -34,6 +34,7 @@ def init_args():
parser = argparse.ArgumentParser()
# params for prediction engine
parser.add_argument("--use_gpu", type=str2bool, default=True)
+ parser.add_argument("--use_xpu", type=str2bool, default=False)
parser.add_argument("--ir_optim", type=str2bool, default=True)
parser.add_argument("--use_tensorrt", type=str2bool, default=False)
parser.add_argument("--min_subgraph_size", type=int, default=15)
@@ -201,7 +202,8 @@ def create_predictor(args, mode, logger):
workspace_size=1 << 30,
precision_mode=precision,
max_batch_size=args.max_batch_size,
- min_subgraph_size=args.min_subgraph_size)
+ min_subgraph_size=args.min_subgraph_size,
+ use_calib_mode=False)
# skip the minmum trt subgraph
use_dynamic_shape = True
if mode == "det":
@@ -275,6 +277,7 @@ def create_predictor(args, mode, logger):
min_input_shape = {"x": [1, 3, imgH, 10]}
max_input_shape = {"x": [args.rec_batch_num, 3, imgH, 2304]}
opt_input_shape = {"x": [args.rec_batch_num, 3, imgH, 320]}
+ config.exp_disable_tensorrt_ops(["transpose2"])
elif mode == "cls":
min_input_shape = {"x": [1, 3, 48, 10]}
max_input_shape = {"x": [args.rec_batch_num, 3, 48, 1024]}
@@ -285,6 +288,8 @@ def create_predictor(args, mode, logger):
config.set_trt_dynamic_shape_info(
min_input_shape, max_input_shape, opt_input_shape)
+ elif args.use_xpu:
+ config.enable_xpu(10 * 1024 * 1024)
else:
config.disable_gpu()
if hasattr(args, "cpu_threads"):
diff --git a/tools/infer_rec.py b/tools/infer_rec.py
index 193e24a4de..a08fa25b46 100755
--- a/tools/infer_rec.py
+++ b/tools/infer_rec.py
@@ -157,7 +157,7 @@ def main():
if info is not None:
logger.info("\t result: {}".format(info))
- fout.write(file + "\t" + info)
+ fout.write(file + "\t" + info + "\n")
logger.info("success!")
diff --git a/tools/program.py b/tools/program.py
index 7c02dc0149..aa3ba82c44 100755
--- a/tools/program.py
+++ b/tools/program.py
@@ -112,20 +112,25 @@ def merge_config(config, opts):
return config
-def check_gpu(use_gpu):
+def check_device(use_gpu, use_xpu=False):
"""
Log error and exit when set use_gpu=true in paddlepaddle
cpu version.
"""
- err = "Config use_gpu cannot be set as true while you are " \
- "using paddlepaddle cpu version ! \nPlease try: \n" \
- "\t1. Install paddlepaddle-gpu to run model on GPU \n" \
- "\t2. Set use_gpu as false in config file to run " \
+ err = "Config {} cannot be set as true while your paddle " \
+ "is not compiled with {} ! \nPlease try: \n" \
+ "\t1. Install paddlepaddle to run model on {} \n" \
+ "\t2. Set {} as false in config file to run " \
"model on CPU"
try:
+ if use_gpu and use_xpu:
+ print("use_xpu and use_gpu can not both be ture.")
if use_gpu and not paddle.is_compiled_with_cuda():
- print(err)
+ print(err.format("use_gpu", "cuda", "gpu", "use_gpu"))
+ sys.exit(1)
+ if use_xpu and not paddle.device.is_compiled_with_xpu():
+ print(err.format("use_xpu", "xpu", "xpu", "use_xpu"))
sys.exit(1)
except Exception as e:
pass
@@ -250,6 +255,8 @@ def train(config,
with paddle.amp.auto_cast():
if model_type == 'table' or extra_input:
preds = model(images, data=batch[1:])
+ elif model_type in ["kie", 'vqa']:
+ preds = model(batch)
else:
preds = model(images)
else:
@@ -302,7 +309,8 @@ def train(config,
train_stats.update(stats)
if log_writer is not None and dist.get_rank() == 0:
- log_writer.log_metrics(metrics=train_stats.get(), prefix="TRAIN", step=global_step)
+ log_writer.log_metrics(
+ metrics=train_stats.get(), prefix="TRAIN", step=global_step)
if dist.get_rank() == 0 and (
(global_step > 0 and global_step % print_batch_step == 0) or
@@ -349,7 +357,8 @@ def train(config,
# logger metric
if log_writer is not None:
- log_writer.log_metrics(metrics=cur_metric, prefix="EVAL", step=global_step)
+ log_writer.log_metrics(
+ metrics=cur_metric, prefix="EVAL", step=global_step)
if cur_metric[main_indicator] >= best_model_dict[
main_indicator]:
@@ -372,11 +381,18 @@ def train(config,
logger.info(best_str)
# logger best metric
if log_writer is not None:
- log_writer.log_metrics(metrics={
- "best_{}".format(main_indicator): best_model_dict[main_indicator]
- }, prefix="EVAL", step=global_step)
-
- log_writer.log_model(is_best=True, prefix="best_accuracy", metadata=best_model_dict)
+ log_writer.log_metrics(
+ metrics={
+ "best_{}".format(main_indicator):
+ best_model_dict[main_indicator]
+ },
+ prefix="EVAL",
+ step=global_step)
+
+ log_writer.log_model(
+ is_best=True,
+ prefix="best_accuracy",
+ metadata=best_model_dict)
reader_start = time.time()
if dist.get_rank() == 0:
@@ -408,7 +424,8 @@ def train(config,
epoch=epoch,
global_step=global_step)
if log_writer is not None:
- log_writer.log_model(is_best=False, prefix='iter_epoch_{}'.format(epoch))
+ log_writer.log_model(
+ is_best=False, prefix='iter_epoch_{}'.format(epoch))
best_str = 'best metric, {}'.format(', '.join(
['{}: {}'.format(k, v) for k, v in best_model_dict.items()]))
@@ -547,7 +564,7 @@ def preprocess(is_train=False):
# check if set use_gpu=True in paddlepaddle cpu version
use_gpu = config['Global']['use_gpu']
- check_gpu(use_gpu)
+ use_xpu = config['Global'].get('use_xpu', False)
# check if set use_xpu=True in paddlepaddle cpu/gpu version
use_xpu = False
@@ -559,14 +576,17 @@ def preprocess(is_train=False):
assert alg in [
'EAST', 'DB', 'SAST', 'Rosetta', 'CRNN', 'STARNet', 'RARE', 'SRN',
'CLS', 'PGNet', 'Distillation', 'NRTR', 'TableAttn', 'SAR', 'PSE',
- 'SEED', 'SDMGR', 'LayoutXLM', 'LayoutLM', 'PREN', 'FCE', 'SVTR'
+ 'SEED', 'SDMGR', 'LayoutXLM', 'LayoutLM', 'PREN', 'FCE', 'SVTR',
+ 'ViTSTR', 'ABINet'
]
- device = 'cpu'
- if use_gpu:
- device = 'gpu:{}'.format(dist.ParallelEnv().dev_id)
if use_xpu:
- device = 'xpu'
+ device = 'xpu:{0}'.format(os.getenv('FLAGS_selected_xpus', 0))
+ else:
+ device = 'gpu:{}'.format(dist.ParallelEnv()
+ .dev_id) if use_gpu else 'cpu'
+ check_device(use_gpu, use_xpu)
+
device = paddle.set_device(device)
config['Global']['distributed'] = dist.get_world_size() != 1
@@ -578,7 +598,8 @@ def preprocess(is_train=False):
vdl_writer_path = '{}/vdl/'.format(save_model_dir)
log_writer = VDLLogger(save_model_dir)
loggers.append(log_writer)
- if ('use_wandb' in config['Global'] and config['Global']['use_wandb']) or 'wandb' in config:
+ if ('use_wandb' in config['Global'] and
+ config['Global']['use_wandb']) or 'wandb' in config:
save_dir = config['Global']['save_model_dir']
wandb_writer_path = "{}/wandb".format(save_dir)
if "wandb" in config:
diff --git a/tools/train.py b/tools/train.py
index 42aba548d6..b7c25e3423 100755
--- a/tools/train.py
+++ b/tools/train.py
@@ -35,6 +35,7 @@ from ppocr.postprocess import build_post_process
from ppocr.metrics import build_metric
from ppocr.utils.save_load import load_model
from ppocr.utils.utility import set_seed
+from ppocr.modeling.architectures import apply_to_static
import tools.program as program
dist.get_world_size()
@@ -121,6 +122,8 @@ def main(config, device, logger, vdl_writer):
if config['Global']['distributed']:
model = paddle.DataParallel(model)
+ model = apply_to_static(model, config, logger)
+
# build loss
loss_class = build_loss(config['Loss'])
+
+
+目前光功率计缺少将数据直接输出的功能，需要人工读数。这一项工作单调重复，如果可以使用机器替代人工，将节约大量成本。针对上述问题，希望通过摄像头拍照->智能读数的方式高效地完成此任务。
+
+为实现智能读数，通常会采取文本检测+文本识别的方案：
+
+第一步，使用文本检测模型定位出光功率计中的数字部分；
+
+第二步，使用文本识别模型获得准确的数字和单位信息。
+
+本项目主要介绍如何完成第二步文本识别部分，包括：真实评估集的建立、训练数据的合成、基于 PP-OCRv3 和 SVTR_Tiny 两个模型进行训练，以及评估和推理。
+
+本项目难点如下：
+
+- 光功率计数码管字符数据较少，难以获取。
+- 数码管中小数点占像素较少，容易漏识别。
+
+针对以上问题， 本例选用 PP-OCRv3 和 SVTR_Tiny 两个高精度模型训练，同时提供了真实数据挖掘案例和数据合成案例。基于 PP-OCRv3 模型，在构建的真实评估集上精度从 52% 提升至 72%，SVTR_Tiny 模型精度可达到 78.9%。
+
+aistudio项目链接: [光功率计数码管字符识别](https://aistudio.baidu.com/aistudio/projectdetail/4049044?contributionType=1)
+
+## 2. PaddleOCR 快速使用
+
+PaddleOCR 旨在打造一套丰富、领先、且实用的OCR工具库，助力开发者训练出更好的模型，并应用落地。
+
+
+
+
+官方提供了适用于通用场景的高精轻量模型，首先使用官方提供的 PP-OCRv3 模型预测图片，验证下当前模型在光功率计场景上的效果。
+
+- 准备环境
+
+```
+python3 -m pip install -U pip
+python3 -m pip install paddleocr
+```
+
+
+- 测试效果
+
+测试图：
+
+
+
+
+```
+paddleocr --lang=ch --det=Fase --image_dir=data
+```
+
+得到如下测试结果：
+
+```
+('.7000', 0.6885431408882141)
+```
+
+发现数字识别较准，然而对负号和小数点识别不准确。 由于PP-OCRv3的训练数据大多为通用场景数据，在特定的场景上效果可能不够好。因此需要基于场景数据进行微调。
+
+下面就主要介绍如何在光功率计（数码管）场景上微调训练。
+
+
+## 3. 开始训练
+
+### 3.1 数据准备
+
+特定的工业场景往往很难获取开源的真实数据集，光功率计也是如此。在实际工业场景中，可以通过摄像头采集的方法收集大量真实数据，本例中重点介绍数据合成方法和真实数据挖掘方法，如何利用有限的数据优化模型精度。
+
+数据集分为两个部分：合成数据，真实数据， 其中合成数据由 text_renderer 工具批量生成得到， 真实数据通过爬虫等方式在百度图片中搜索并使用 PPOCRLabel 标注得到。
+
+
+- 合成数据
+
+本例中数据合成工具使用的是 [text_renderer](https://github.com/Sanster/text_renderer)， 该工具可以合成用于文本识别训练的文本行数据:
+
+
+
+
+
+
+```
+export https_proxy=http://172.19.57.45:3128
+git clone https://github.com/oh-my-ocr/text_renderer
+```
+
+```
+import os
+python3 setup.py develop
+python3 -m pip install -r docker/requirements.txt
+python3 main.py \
+ --config example_data/example.py \
+ --dataset img \
+ --num_processes 2 \
+ --log_period 10
+```
+
+给定字体和语料，就可以合成较为丰富样式的文本行数据。 光功率计识别场景，目标是正确识别数码管文本，因此需要收集部分数码管字体，训练语料，用于合成文本识别数据。
+
+将收集好的语料存放在 example_data 路径下：
+
+```
+ln -s ./fonts/DS* text_renderer/example_data/font/
+ln -s ./corpus/digital.txt text_renderer/example_data/text/
+```
+
+修改 text_renderer/example_data/font_list/font_list.txt ,选择需要的字体开始合成：
+
+```
+python3 main.py \
+ --config example_data/digital_example.py \
+ --dataset img \
+ --num_processes 2 \
+ --log_period 10
+```
+
+合成图片会被存在目录 text_renderer/example_data/digital/chn_data 下
+
+查看合成的数据样例：
+
+
+
+
+- 真实数据挖掘
+
+模型训练需要使用真实数据作为评价指标，否则很容易过拟合到简单的合成数据中。没有开源数据的情况下，可以利用部分无标注数据+标注工具获得真实数据。
+
+
+1. 数据搜集
+
+使用[爬虫工具](https://github.com/Joeclinton1/google-images-download.git)获得无标注数据
+
+2. [PPOCRLabel](https://github.com/PaddlePaddle/PaddleOCR/tree/release/2.5/PPOCRLabel) 完成半自动标注
+
+PPOCRLabel是一款适用于OCR领域的半自动化图形标注工具，内置PP-OCR模型对数据自动标注和重新识别。使用Python3和PyQT5编写，支持矩形框标注、表格标注、不规则文本标注、关键信息标注模式，导出格式可直接用于PaddleOCR检测和识别模型的训练。
+
+
+
+
+收集完数据后就可以进行分配了，验证集中一般都是真实数据，训练集中包含合成数据+真实数据。本例中标注了155张图片，其中训练集和验证集的数目为100和55。
+
+
+最终 `data` 文件夹应包含以下几部分：
+
+```
+|-data
+ |- synth_train.txt
+ |- real_train.txt
+ |- real_eval.txt
+ |- synthetic_data
+ |- word_001.png
+ |- word_002.jpg
+ |- word_003.jpg
+ | ...
+ |- real_data
+ |- word_001.png
+ |- word_002.jpg
+ |- word_003.jpg
+ | ...
+ ...
+```
+
+### 3.2 模型选择
+
+本案例提供了2种文本识别模型：PP-OCRv3 识别模型 和 SVTR_Tiny：
+
+[PP-OCRv3 识别模型](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.5/doc/doc_ch/PP-OCRv3_introduction.md)：PP-OCRv3的识别模块是基于文本识别算法SVTR优化。SVTR不再采用RNN结构，通过引入Transformers结构更加有效地挖掘文本行图像的上下文信息，从而提升文本识别能力。并进行了一系列结构改进加速模型预测。
+
+[SVTR_Tiny](https://arxiv.org/abs/2205.00159):SVTR提出了一种用于场景文本识别的单视觉模型，该模型在patch-wise image tokenization框架内，完全摒弃了序列建模，在精度具有竞争力的前提下，模型参数量更少，速度更快。
+
+以上两个策略在自建中文数据集上的精度和速度对比如下：
+
+| ID | 策略 | 模型大小 | 精度 | 预测耗时（CPU + MKLDNN)|
+|-----|-----|--------|----| --- |
+| 01 | PP-OCRv2 | 8M | 74.8% | 8.54ms |
+| 02 | SVTR_Tiny | 21M | 80.1% | 97ms |
+| 03 | SVTR_LCNet(h32) | 12M | 71.9% | 6.6ms |
+| 04 | SVTR_LCNet(h48) | 12M | 73.98% | 7.6ms |
+| 05 | + GTC | 12M | 75.8% | 7.6ms |
+| 06 | + TextConAug | 12M | 76.3% | 7.6ms |
+| 07 | + TextRotNet | 12M | 76.9% | 7.6ms |
+| 08 | + UDML | 12M | 78.4% | 7.6ms |
+| 09 | + UIM | 12M | 79.4% | 7.6ms |
+
+
+### 3.3 开始训练
+
+首先下载 PaddleOCR 代码库
+
+```
+git clone -b release/2.5 https://github.com/PaddlePaddle/PaddleOCR.git
+```
+
+PaddleOCR提供了训练脚本、评估脚本和预测脚本，本节将以 PP-OCRv3 中文识别模型为例：
+
+**Step1：下载预训练模型**
+
+首先下载 pretrain model，您可以下载训练好的模型在自定义数据上进行finetune
+
+```
+cd PaddleOCR/
+# 下载PP-OCRv3 中文预训练模型
+wget -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_train.tar
+# 解压模型参数
+cd pretrain_models
+tar -xf ch_PP-OCRv3_rec_train.tar && rm -rf ch_PP-OCRv3_rec_train.tar
+```
+
+**Step2：自定义字典文件**
+
+接下来需要提供一个字典（{word_dict_name}.txt），使模型在训练时，可以将所有出现的字符映射为字典的索引。
+
+因此字典需要包含所有希望被正确识别的字符，{word_dict_name}.txt需要写成如下格式，并以 `utf-8` 编码格式保存：
+
+```
+0
+1
+2
+3
+4
+5
+6
+7
+8
+9
+-
+.
+```
+
+word_dict.txt 每行有一个单字，将字符与数字索引映射在一起，“3.14” 将被映射成 [3, 11, 1, 4]
+
+* 内置字典
+
+PaddleOCR内置了一部分字典，可以按需使用。
+
+`ppocr/utils/ppocr_keys_v1.txt` 是一个包含6623个字符的中文字典
+
+`ppocr/utils/ic15_dict.txt` 是一个包含36个字符的英文字典
+
+* 自定义字典
+
+内置字典面向通用场景，具体的工业场景中，可能需要识别特殊字符，或者只需识别某几个字符，此时自定义字典会更提升模型精度。例如在光功率计场景中，需要识别数字和单位。
+
+遍历真实数据标签中的字符，制作字典`digital_dict.txt`如下所示：
+
+```
+-
+.
+0
+1
+2
+3
+4
+5
+6
+7
+8
+9
+B
+E
+F
+H
+L
+N
+T
+W
+d
+k
+m
+n
+o
+z
+```
+
+
+
+
+**Step3：修改配置文件**
+
+为了更好的使用预训练模型，训练推荐使用[ch_PP-OCRv3_rec_distillation.yml](../../configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml)配置文件，并参考下列说明修改配置文件：
+
+以 `ch_PP-OCRv3_rec_distillation.yml` 为例：
+```
+Global:
+ ...
+ # 添加自定义字典，如修改字典请将路径指向新字典
+ character_dict_path: ppocr/utils/dict/digital_dict.txt
+ ...
+ # 识别空格
+ use_space_char: True
+
+
+Optimizer:
+ ...
+ # 添加学习率衰减策略
+ lr:
+ name: Cosine
+ learning_rate: 0.001
+ ...
+
+...
+
+Train:
+ dataset:
+ # 数据集格式，支持LMDBDataSet以及SimpleDataSet
+ name: SimpleDataSet
+ # 数据集路径
+ data_dir: ./data/
+ # 训练集标签文件
+ label_file_list:
+ - ./train_data/digital_img/digital_train.txt #11w
+ - ./train_data/digital_img/real_train.txt #100
+ - ./train_data/digital_img/dbm_img/dbm.txt #3w
+ ratio_list:
+ - 0.3
+ - 1.0
+ - 1.0
+ transforms:
+ ...
+ - RecResizeImg:
+ # 修改 image_shape 以适应长文本
+ image_shape: [3, 48, 320]
+ ...
+ loader:
+ ...
+ # 单卡训练的batch_size
+ batch_size_per_card: 256
+ ...
+
+Eval:
+ dataset:
+ # 数据集格式，支持LMDBDataSet以及SimpleDataSet
+ name: SimpleDataSet
+ # 数据集路径
+ data_dir: ./data
+ # 验证集标签文件
+ label_file_list:
+ - ./train_data/digital_img/real_val.txt
+ transforms:
+ ...
+ - RecResizeImg:
+ # 修改 image_shape 以适应长文本
+ image_shape: [3, 48, 320]
+ ...
+ loader:
+ # 单卡验证的batch_size
+ batch_size_per_card: 256
+ ...
+```
+**注意，训练/预测/评估时的配置文件请务必与训练一致。**
+
+**Step4：启动训练**
+
+*如果您安装的是cpu版本，请将配置文件中的 `use_gpu` 字段修改为false*
+
+```
+# GPU训练 支持单卡，多卡训练
+# 训练数码管数据 训练日志会自动保存为 "{save_model_dir}" 下的train.log
+
+#单卡训练（训练周期长，不建议）
+python3 tools/train.py -c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml -o Global.pretrained_model=./pretrain_models/ch_PP-OCRv3_rec_train/best_accuracy
+
+#多卡训练，通过--gpus参数指定卡号
+python3 -m paddle.distributed.launch --gpus '0,1,2,3' tools/train.py -c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml -o Global.pretrained_model=./pretrain_models/en_PP-OCRv3_rec_train/best_accuracy
+```
+
+
+PaddleOCR支持训练和评估交替进行, 可以在 `configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml` 中修改 `eval_batch_step` 设置评估频率，默认每500个iter评估一次。评估过程中默认将最佳acc模型，保存为 `output/ch_PP-OCRv3_rec_distill/best_accuracy` 。
+
+如果验证集很大，测试将会比较耗时，建议减少评估次数，或训练完再进行评估。
+
+### SVTR_Tiny 训练
+
+SVTR_Tiny 训练步骤与上面一致，SVTR支持的配置和模型训练权重可以参考[算法介绍文档](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.5/doc/doc_ch/algorithm_rec_svtr.md)
+
+**Step1：下载预训练模型**
+
+```
+# 下载 SVTR_Tiny 中文识别预训练模型和配置文件
+wget https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/rec_svtr_tiny_none_ctc_ch_train.tar
+# 解压模型参数
+tar -xf rec_svtr_tiny_none_ctc_ch_train.tar && rm -rf rec_svtr_tiny_none_ctc_ch_train.tar
+```
+**Step2：自定义字典文件**
+
+字典依然使用自定义的 digital_dict.txt
+
+**Step3：修改配置文件**
+
+配置文件中对应修改字典路径和数据路径
+
+**Step4：启动训练**
+
+```
+## 单卡训练
+python tools/train.py -c rec_svtr_tiny_none_ctc_ch_train/rec_svtr_tiny_6local_6global_stn_ch.yml \
+ -o Global.pretrained_model=./rec_svtr_tiny_none_ctc_ch_train/best_accuracy
+```
+
+### 3.4 验证效果
+
+如需获取已训练模型，请扫码填写问卷，加入PaddleOCR官方交流群获取全部OCR垂类模型下载链接、《动手学OCR》电子书等全套OCR学习资料🎁
+
- 
+ 
+ 
- 
+ 
+ 
- 
+ 
+ 