Skip to content

P
PaddleOCR

s920243400 / PaddleOCR
与 Fork 源项目一致

Fork自 PaddlePaddle / PaddleOCR

通知 1
Star 1
Fork 0
P
PaddleOCR
提交 5016894e 编写于 作者: L LDOUBLEV
浏览文件
操作

fix conflict

上级 16bafede a3dfc10b
显示空白变更内容
内联 并排
Showing with 4964 addition and 1614 deletion
applications/PCB字符识别/PCB字符识别.md 0 → 100644
浏览文件 @ 5016894e
# 基于PP-OCRv3的PCB字符识别
- [1. 项目介绍](#1-项目介绍)
- [2. 安装说明](#2-安装说明)
- [3. 数据准备](#3-数据准备)
- [4. 文本检测](#4-文本检测)
- [4.1 预训练模型直接评估](#41-预训练模型直接评估)
- [4.2 预训练模型+验证集padding直接评估](#42-预训练模型验证集padding直接评估)
- [4.3 预训练模型+fine-tune](#43-预训练模型fine-tune)
- [5. 文本识别](#5-文本识别)
- [5.1 预训练模型直接评估](#51-预训练模型直接评估)
- [5.2 三种fine-tune方案](#52-三种fine-tune方案)
- [6. 模型导出](#6-模型导出)
- [7. 端对端评测](#7-端对端评测)
- [8. Jetson部署](#8-Jetson部署)
- [9. 总结](#9-总结)
- [更多资源](#更多资源)
# 1. 项目介绍
印刷电路板(PCB)是电子产品中的核心器件，对于板件质量的测试与监控是生产中必不可少的环节。在一些场景中，通过PCB中信号灯颜色和文字组合可以定位PCB局部模块质量问题，PCB文字识别中存在如下难点：
- 裁剪出的PCB图片宽高比例较小
- 文字区域整体面积也较小
- 包含垂直、水平多种方向文本
针对本场景，PaddleOCR基于全新的PP-OCRv3通过合成数据、微调以及其他场景适配方法完成小字符文本识别任务，满足企业上线要求。PCB检测、识别效果如 **图1** 所示：
<div align=center><img src='https://ai-studio-static-online.cdn.bcebos.com/95d8e95bf1ab476987f2519c0f8f0c60a0cdc2c444804ed6ab08f2f7ab054880', width='500'></div>
<div align=center>图1 PCB检测识别效果</div>
注：欢迎在AIStudio领取免费算力体验线上实训，项目链接: [基于PP-OCRv3实现PCB字符识别](https://aistudio.baidu.com/aistudio/projectdetail/4008973)
# 2. 安装说明
下载PaddleOCR源码，安装依赖环境。
```python
# 如仍需安装or安装更新，可以执行以下步骤
git clone https://github.com/PaddlePaddle/PaddleOCR.git
# git clone https://gitee.com/PaddlePaddle/PaddleOCR
```
```python
# 安装依赖包
pip install -r /home/aistudio/PaddleOCR/requirements.txt
```
# 3. 数据准备
我们通过图片合成工具生成 **图2** 所示的PCB图片，整图只有高25、宽150左右、文字区域高9、宽45左右，包含垂直和水平2种方向的文本：
<div align=center><img src="https://ai-studio-static-online.cdn.bcebos.com/bb7a345687814a3d83a29790f2a2b7d081495b3a920b43988c93da6039cad653" width="1000" ></div>
<div align=center>图2 数据集示例</div>
暂时不开源生成的PCB数据集，但是通过更换背景，通过如下代码生成数据即可：
```
cd gen_data
python3 gen.py --num_img=10
```
生成图片参数解释：
```
num_img：生成图片数量
font_min_size、font_max_size：字体最大、最小尺寸
bg_path：文字区域背景存放路径
det_bg_path：整图背景存放路径
fonts_path：字体路径
corpus_path：语料路径
output_dir：生成图片存储路径
```
这里生成 **100张** 相同尺寸和文本的图片，如 **图3** 所示，方便大家跑通实验。通过如下代码解压数据集：
<div align=center><img src="https://ai-studio-static-online.cdn.bcebos.com/3277b750159f4b68b2b58506bfec9005d49aeb5fb1d9411e83f96f9ff7eb66a5" width="1000" ></div>
<div align=center>图3 案例提供数据集示例</div>
```python
tar xf ./data/data148165/dataset.tar -C ./
```
在生成数据集的时需要生成检测和识别训练需求的格式：
- **文本检测**
标注文件格式如下，中间用'\t'分隔：
```
" 图像文件名 json.dumps编码的图像标注信息"
ch4_test_images/img_61.jpg [{"transcription": "MASA", "points": [[310, 104], [416, 141], [418, 216], [312, 179]]}, {...}]
```
json.dumps编码前的图像标注信息是包含多个字典的list，字典中的 `points` 表示文本框的四个点的坐标(x, y)，从左上角的点开始顺时针排列。 `transcription` 表示当前文本框的文字，***当其内容为“###”时，表示该文本框无效，在训练时会跳过。***
- **文本识别**
标注文件的格式如下， txt文件中默认请将图片路径和图片标签用'\t'分割，如用其他方式分割将造成训练报错。
```
" 图像文件名 图像标注信息 "
train_data/rec/train/word_001.jpg 简单可依赖
train_data/rec/train/word_002.jpg 用科技让复杂的世界更简单
...
```
# 4. 文本检测
选用飞桨OCR开发套件[PaddleOCR](https://github.com/PaddlePaddle/PaddleOCR)中的PP-OCRv3模型进行文本检测和识别。针对检测模型和识别模型，进行了共计9个方面的升级：
- PP-OCRv3检测模型对PP-OCRv2中的CML协同互学习文本检测蒸馏策略进行了升级，分别针对教师模型和学生模型进行进一步效果优化。其中，在对教师模型优化时，提出了大感受野的PAN结构LK-PAN和引入了DML蒸馏策略；在对学生模型优化时，提出了残差注意力机制的FPN结构RSE-FPN。
- 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。PP-OCRv3通过轻量级文本识别网络SVTR_LCNet、Attention损失指导CTC损失训练策略、挖掘文字上下文信息的数据增广策略TextConAug、TextRotNet自监督预训练模型、UDML联合互学习策略、UIM无标注数据挖掘方案，6个方面进行模型加速和效果提升。
更多细节请参考PP-OCRv3[技术报告](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.5/doc/doc_ch/PP-OCRv3_introduction.md)。
我们使用 **3种方案** 进行检测模型的训练、评估：
- **PP-OCRv3英文超轻量检测预训练模型直接评估**
- PP-OCRv3英文超轻量检测预训练模型 + **验证集padding**直接评估
- PP-OCRv3英文超轻量检测预训练模型 + **fine-tune**
## **4.1 预训练模型直接评估**
我们首先通过PaddleOCR提供的预训练模型在验证集上进行评估，如果评估指标能满足效果，可以直接使用预训练模型，不再需要训练。
使用预训练模型直接评估步骤如下：
**1）下载预训练模型**
PaddleOCR已经提供了PP-OCR系列模型，部分模型展示如下表所示：
| 模型简介 | 模型名称 | 推荐场景 | 检测模型 | 方向分类器 | 识别模型 |
| ------------------------------------- | ----------------------- | --------------- | ------------------------------------------------------------ | ------------------------------------------------------------ | ------------------------------------------------------------ |
| 中英文超轻量PP-OCRv3模型（16.2M） | ch_PP-OCRv3_xx | 移动端&服务器端 | [推理模型](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_distill_train.tar) | [推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_train.tar) | [推理模型](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_train.tar) |
| 英文超轻量PP-OCRv3模型（13.4M） | en_PP-OCRv3_xx | 移动端&服务器端 | [推理模型](https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_det_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_det_distill_train.tar) | [推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_train.tar) | [推理模型](https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_rec_train.tar) |
| 中英文超轻量PP-OCRv2模型（13.0M） | ch_PP-OCRv2_xx | 移动端&服务器端 | [推理模型](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_distill_train.tar) | [推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_train.tar) | [推理模型](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_train.tar) |
| 中英文超轻量PP-OCR mobile模型（9.4M） | ch_ppocr_mobile_v2.0_xx | 移动端&服务器端 | [推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_train.tar) | [推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_train.tar) | [推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_pre.tar) |
| 中英文通用PP-OCR server模型（143.4M） | ch_ppocr_server_v2.0_xx | 服务器端 | [推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_train.tar) | [推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_train.tar) | [推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_rec_infer.tar) / [预训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_rec_pre.tar) |
更多模型下载（包括多语言），可以参[考PP-OCR系列模型下载](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.5/doc/doc_ch/models_list.md)
这里我们使用PP-OCRv3英文超轻量检测模型，下载并解压预训练模型：
```python
# 如果更换其他模型，更新下载链接和解压指令就可以
cd /home/aistudio/PaddleOCR
mkdir pretrain_models
cd pretrain_models
# 下载英文预训练模型
wget https://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_det_distill_train.tar
tar xf en_PP-OCRv3_det_distill_train.tar && rm -rf en_PP-OCRv3_det_distill_train.tar
%cd ..
```
**模型评估**
首先修改配置文件`configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml`中的以下字段：
```
Eval.dataset.data_dir：指向验证集图片存放目录,'/home/aistudio/dataset'
Eval.dataset.label_file_list：指向验证集标注文件,'/home/aistudio/dataset/det_gt_val.txt'
Eval.dataset.transforms.DetResizeForTest: 尺寸
limit_side_len: 48
limit_type: 'min'
```
然后在验证集上进行评估，具体代码如下：
```python
cd /home/aistudio/PaddleOCR
python tools/eval.py \
-c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml \
-o Global.checkpoints="./pretrain_models/en_PP-OCRv3_det_distill_train/best_accuracy"
```
## **4.2 预训练模型+验证集padding直接评估**
考虑到PCB图片比较小，宽度只有25左右、高度只有140-170左右，我们在原图的基础上进行padding，再进行检测评估，padding前后效果对比如 **图4** 所示：
<div align=center><img src='https://ai-studio-static-online.cdn.bcebos.com/e61e6ba685534eda992cea30a63a9c461646040ffd0c4d208a5eebb85897dcf7' width='600'></div>
<div align=center>图4 padding前后对比图</div>
将图片都padding到300*300大小，因为坐标信息发生了变化，我们同时要修改标注文件，在`/home/aistudio/dataset`目录里也提供了padding之后的图片，大家也可以尝试训练和评估：
同上，我们需要修改配置文件`configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml`中的以下字段：
```
Eval.dataset.data_dir：指向验证集图片存放目录,'/home/aistudio/dataset'
Eval.dataset.label_file_list：指向验证集标注文件,/home/aistudio/dataset/det_gt_padding_val.txt
Eval.dataset.transforms.DetResizeForTest: 尺寸
limit_side_len: 1100
limit_type: 'min'
```
然后执行评估代码
```python
cd /home/aistudio/PaddleOCR
python tools/eval.py \
-c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml \
-o Global.checkpoints="./pretrain_models/en_PP-OCRv3_det_distill_train/best_accuracy"
```
## **4.3 预训练模型+fine-tune**
基于预训练模型，在生成的1500图片上进行fine-tune训练和评估，其中train数据1200张，val数据300张，修改配置文件`configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_student.yml`中的以下字段：
```
Global.epoch_num: 这里设置为1，方便快速跑通，实际中根据数据量调整该值
Global.save_model_dir：模型保存路径
Global.pretrained_model：指向预训练模型路径，'./pretrain_models/en_PP-OCRv3_det_distill_train/student.pdparams'
Optimizer.lr.learning_rate：调整学习率，本实验设置为0.0005
Train.dataset.data_dir：指向训练集图片存放目录,'/home/aistudio/dataset'
Train.dataset.label_file_list：指向训练集标注文件,'/home/aistudio/dataset/det_gt_train.txt'
Train.dataset.transforms.EastRandomCropData.size：训练尺寸改为[480,64]
Eval.dataset.data_dir：指向验证集图片存放目录,'/home/aistudio/dataset/'
Eval.dataset.label_file_list：指向验证集标注文件,'/home/aistudio/dataset/det_gt_val.txt'
Eval.dataset.transforms.DetResizeForTest：评估尺寸，添加如下参数
limit_side_len: 64
limit_type:'min'
```
执行下面命令启动训练：
```python
cd /home/aistudio/PaddleOCR/
python tools/train.py \
-c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_student.yml
```
**模型评估**
使用训练好的模型进行评估，更新模型路径`Global.checkpoints`:
```python
cd /home/aistudio/PaddleOCR/
python3 tools/eval.py \
-c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_student.yml \
-o Global.checkpoints="./output/ch_PP-OCR_V3_det/latest"
```
使用训练好的模型进行评估，指标如下所示：
| 序号 | 方案 | hmean | 效果提升 | 实验分析 |
| -------- | -------- | -------- | -------- | -------- |
| 1 | PP-OCRv3英文超轻量检测预训练模型 | 64.64% | - | 提供的预训练模型具有泛化能力 |
| 2 | PP-OCRv3英文超轻量检测预训练模型 + 验证集padding | 72.13% |+7.5% | padding可以提升尺寸较小图片的检测效果|
| 3 | PP-OCRv3英文超轻量检测预训练模型 + fine-tune | 100% | +27.9% | fine-tune会提升垂类场景效果 |
```
注：上述实验结果均是在1500张图片（1200张训练集，300张测试集）上训练、评估的得到，AIstudio只提供了100张数据，所以指标有所差异属于正常，只要策略有效、规律相同即可。
```
# 5. 文本识别
我们分别使用如下4种方案进行训练、评估：
- **方案1**：**PP-OCRv3中英文超轻量识别预训练模型直接评估**
- **方案2**：PP-OCRv3中英文超轻量检测预训练模型 + **fine-tune**
- **方案3**：PP-OCRv3中英文超轻量检测预训练模型 + fine-tune + **公开通用识别数据集**
- **方案4**：PP-OCRv3中英文超轻量检测预训练模型 + fine-tune + **增加PCB图像数量**
## **5.1 预训练模型直接评估**
同检测模型，我们首先使用PaddleOCR提供的识别预训练模型在PCB验证集上进行评估。
使用预训练模型直接评估步骤如下：
**1）下载预训练模型**
我们使用PP-OCRv3中英文超轻量文本识别模型，下载并解压预训练模型：
```python
# 如果更换其他模型，更新下载链接和解压指令就可以
cd /home/aistudio/PaddleOCR/pretrain_models/
wget https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_train.tar
tar xf ch_PP-OCRv3_rec_train.tar && rm -rf ch_PP-OCRv3_rec_train.tar
cd ..
```
**模型评估**
首先修改配置文件`configs/det/ch_PP-OCRv3/ch_PP-OCRv2_rec_distillation.yml`中的以下字段：
```
Metric.ignore_space: True：忽略空格
Eval.dataset.data_dir：指向验证集图片存放目录,'/home/aistudio/dataset'
Eval.dataset.label_file_list：指向验证集标注文件,'/home/aistudio/dataset/rec_gt_val.txt'
```
我们使用下载的预训练模型进行评估：
```python
cd /home/aistudio/PaddleOCR
python3 tools/eval.py \
-c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml \
-o Global.checkpoints=pretrain_models/ch_PP-OCRv3_rec_train/best_accuracy
```
## **5.2 三种fine-tune方案**
方案2、3、4训练和评估方式是相同的，因此在我们了解每个技术方案之后，再具体看修改哪些参数是相同，哪些是不同的。
**方案介绍：**
1） **方案2**：预训练模型 + **fine-tune**
- 在预训练模型的基础上进行fine-tune，使用1500张PCB进行训练和评估，其中训练集1200张，验证集300张。
2） **方案3**：预训练模型 + fine-tune + **公开通用识别数据集**
- 当识别数据比较少的情况，可以考虑添加公开通用识别数据集。在方案2的基础上，添加公开通用识别数据集，如lsvt、rctw等。
3）**方案4**：预训练模型 + fine-tune + **增加PCB图像数量**
- 如果能够获取足够多真实场景，我们可以通过增加数据量提升模型效果。在方案2的基础上，增加PCB的数量到2W张左右。
**参数修改：**
接着我们看需要修改的参数，以上方案均需要修改配置文件`configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml`的参数，**修改一次即可**：
```
Global.pretrained_model：指向预训练模型路径,'pretrain_models/ch_PP-OCRv3_rec_train/best_accuracy'
Optimizer.lr.values：学习率，本实验设置为0.0005
Train.loader.batch_size_per_card: batch size,默认128，因为数据量小于128，因此我们设置为8，数据量大可以按默认的训练
Eval.loader.batch_size_per_card: batch size,默认128，设置为4
Metric.ignore_space: 忽略空格，本实验设置为True
```
**更换不同的方案**每次需要修改的参数：
```
Global.epoch_num: 这里设置为1，方便快速跑通，实际中根据数据量调整该值
Global.save_model_dir：指向模型保存路径
Train.dataset.data_dir：指向训练集图片存放目录
Train.dataset.label_file_list：指向训练集标注文件
Eval.dataset.data_dir：指向验证集图片存放目录
Eval.dataset.label_file_list：指向验证集标注文件
```
同时**方案3**修改以下参数
```
Eval.dataset.label_file_list：添加公开通用识别数据标注文件
Eval.dataset.ratio_list：数据和公开通用识别数据每次采样比例，按实际修改即可
```
如 **图5** 所示：
<div align=center><img src='https://ai-studio-static-online.cdn.bcebos.com/0fa18b25819042d9bbf3397c3af0e21433b23d52f7a84b0a8681b8e6a308d433' wdith=''></div>
<div align=center>图5 添加公开通用识别数据配置文件示例</div>
我们提取Student模型的参数，在PCB数据集上进行fine-tune，可以参考如下代码：
```python
import paddle
# 加载预训练模型
all_params = paddle.load("./pretrain_models/ch_PP-OCRv3_rec_train/best_accuracy.pdparams")
# 查看权重参数的keys
print(all_params.keys())
# 学生模型的权重提取
s_params = {key[len("student_model."):]: all_params[key] for key in all_params if "student_model." in key}
# 查看学生模型权重参数的keys
print(s_params.keys())
# 保存
paddle.save(s_params, "./pretrain_models/ch_PP-OCRv3_rec_train/student.pdparams")
```
修改参数后，**每个方案**都执行如下命令启动训练：
```python
cd /home/aistudio/PaddleOCR/
python3 tools/train.py -c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml
```
使用训练好的模型进行评估，更新模型路径`Global.checkpoints`：
```python
cd /home/aistudio/PaddleOCR/
python3 tools/eval.py \
-c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml \
-o Global.checkpoints=./output/rec_ppocr_v3/latest
```
所有方案评估指标如下：
| 序号 | 方案 | acc | 效果提升 | 实验分析 |
| -------- | -------- | -------- | -------- | -------- |
| 1 | PP-OCRv3中英文超轻量识别预训练模型直接评估 | 46.67% | - | 提供的预训练模型具有泛化能力 |
| 2 | PP-OCRv3中英文超轻量识别预训练模型 + fine-tune | 42.02% |-4.6% | 在数据量不足的情况，反而比预训练模型效果低(也可以通过调整超参数再试试)|
| 3 | PP-OCRv3中英文超轻量识别预训练模型 + fine-tune + 公开通用识别数据集 | 77% | +30% | 在数据量不足的情况下，可以考虑补充公开数据训练 |
| 4 | PP-OCRv3中英文超轻量识别预训练模型 + fine-tune + 增加PCB图像数量 | 99.99% | +23% | 如果能获取更多数据量的情况，可以通过增加数据量提升效果 |
```
注：上述实验结果均是在1500张图片（1200张训练集，300张测试集）、2W张图片、添加公开通用识别数据集上训练、评估的得到，AIstudio只提供了100张数据，所以指标有所差异属于正常，只要策略有效、规律相同即可。
```
# 6. 模型导出
inference 模型（paddle.jit.save保存的模型） 一般是模型训练，把模型结构和模型参数保存在文件中的固化模型，多用于预测部署场景。 训练过程中保存的模型是checkpoints模型，保存的只有模型的参数，多用于恢复训练等。 与checkpoints模型相比，inference 模型会额外保存模型的结构信息，在预测部署、加速推理上性能优越，灵活方便，适合于实际系统集成。
```python
# 导出检测模型
python3 tools/export_model.py \
-c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_student.yml \
-o Global.pretrained_model="./output/ch_PP-OCR_V3_det/latest" \
Global.save_inference_dir="./inference_model/ch_PP-OCR_V3_det/"
```
因为上述模型只训练了1个epoch，因此我们使用训练最优的模型进行预测，存储在`/home/aistudio/best_models/`目录下，解压即可
```python
cd /home/aistudio/best_models/
wget https://paddleocr.bj.bcebos.com/fanliku/PCB/det_ppocr_v3_en_infer_PCB.tar
tar xf /home/aistudio/best_models/det_ppocr_v3_en_infer_PCB.tar -C /home/aistudio/PaddleOCR/pretrain_models/
```
```python
# 检测模型inference模型预测
cd /home/aistudio/PaddleOCR/
python3 tools/infer/predict_det.py \
--image_dir="/home/aistudio/dataset/imgs/0000.jpg" \
--det_algorithm="DB" \
--det_model_dir="./pretrain_models/det_ppocr_v3_en_infer_PCB/" \
--det_limit_side_len=48 \
--det_limit_type='min' \
--det_db_unclip_ratio=2.5 \
--use_gpu=True
```
结果存储在`inference_results`目录下，检测如下图所示：
<div align=center><img src='https://ai-studio-static-online.cdn.bcebos.com/5939ae15a1f0445aaeec15c68107dbd897740a1ddd284bf8b583bb6242099157' width=''></div>
<div align=center>图6 检测结果</div>
同理，导出识别模型并进行推理。
```python
# 导出识别模型
python3 tools/export_model.py \
-c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml \
-o Global.pretrained_model="./output/rec_ppocr_v3/latest" \
Global.save_inference_dir="./inference_model/rec_ppocr_v3/"
```
同检测模型，识别模型也只训练了1个epoch，因此我们使用训练最优的模型进行预测，存储在`/home/aistudio/best_models/`目录下，解压即可
```python
cd /home/aistudio/best_models/
wget https://paddleocr.bj.bcebos.com/fanliku/PCB/rec_ppocr_v3_ch_infer_PCB.tar
tar xf /home/aistudio/best_models/rec_ppocr_v3_ch_infer_PCB.tar -C /home/aistudio/PaddleOCR/pretrain_models/
```
```python
# 识别模型inference模型预测
cd /home/aistudio/PaddleOCR/
python3 tools/infer/predict_rec.py \
--image_dir="../test_imgs/0000_rec.jpg" \
--rec_model_dir="./pretrain_models/rec_ppocr_v3_ch_infer_PCB" \
--rec_image_shape="3, 48, 320" \
--use_space_char=False \
--use_gpu=True
```
```python
# 检测+识别模型inference模型预测
cd /home/aistudio/PaddleOCR/
python3 tools/infer/predict_system.py \
--image_dir="../test_imgs/0000.jpg" \
--det_model_dir="./pretrain_models/det_ppocr_v3_en_infer_PCB" \
--det_limit_side_len=48 \
--det_limit_type='min' \
--det_db_unclip_ratio=2.5 \
--rec_model_dir="./pretrain_models/rec_ppocr_v3_ch_infer_PCB" \
--rec_image_shape="3, 48, 320" \
--draw_img_save_dir=./det_rec_infer/ \
--use_space_char=False \
--use_angle_cls=False \
--use_gpu=True
```
端到端预测结果存储在`det_res_infer`文件夹内，结果如下图所示：
<div align=center><img src='https://ai-studio-static-online.cdn.bcebos.com/c570f343c29846c792da56ebaca16c50708477514dd048cea8bef37ffa85d03f'></div>
<div align=center>图7 检测+识别结果</div>
# 7. 端对端评测
接下来介绍文本检测+文本识别的端对端指标评估方式。主要分为三步：
1）首先运行`tools/infer/predict_system.py`，将`image_dir`改为需要评估的数据文件家，得到保存的结果:
```python
# 检测+识别模型inference模型预测
python3 tools/infer/predict_system.py \
--image_dir="../dataset/imgs/" \
--det_model_dir="./pretrain_models/det_ppocr_v3_en_infer_PCB" \
--det_limit_side_len=48 \
--det_limit_type='min' \
--det_db_unclip_ratio=2.5 \
--rec_model_dir="./pretrain_models/rec_ppocr_v3_ch_infer_PCB" \
--rec_image_shape="3, 48, 320" \
--draw_img_save_dir=./det_rec_infer/ \
--use_space_char=False \
--use_angle_cls=False \
--use_gpu=True
```
得到保存结果，文本检测识别可视化图保存在`det_rec_infer/`目录下，预测结果保存在`det_rec_infer/system_results.txt`中，格式如下：`0018.jpg [{"transcription": "E295", "points": [[88, 33], [137, 33], [137, 40], [88, 40]]}]`
2）然后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/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 = "/home/aistudio/dataset/det_gt_val.txt"
convert_label(ppocr_label_gt, "gt", "./save_gt_label/")
ppocr_label_gt = "/home/aistudio/PaddleOCR/PCB_result/det_rec_infer/system_results.txt"
convert_label(ppocr_label_gt, "pred", "./save_PPOCRV2_infer/")
```
运行`convert_ppocr_label.py`:
```python
python3 tools/end2end/convert_ppocr_label.py
```
得到如下结果：
```
├── ./save_gt_label/
├── ./save_PPOCRV2_infer/
```
3） 最后，执行端对端评测，运行`tools/end2end/eval_end2end.py`计算端对端指标，运行方式如下：
```python
pip install editdistance
python3 tools/end2end/eval_end2end.py ./save_gt_label/ ./save_PPOCRV2_infer/
```
使用`预训练模型+fine-tune'检测模型`、`预训练模型 + 2W张PCB图片funetune`识别模型，在300张PCB图片上评估得到如下结果，fmeasure为主要关注的指标:
<div align=center><img src='https://ai-studio-static-online.cdn.bcebos.com/37206ea48a244212ae7a821d50d1fd51faf3d7fe97ac47a29f04dfcbb377b019', width='700'></div>
<div align=center>图8 端到端评估指标</div>
```
注: 使用上述命令不能跑出该结果，因为数据集不相同，可以更换为自己训练好的模型，按上述流程运行
```
# 8. Jetson部署
我们只需要以下步骤就可以完成Jetson nano部署模型，简单易操作：
**1、在Jetson nano开发版上环境准备：**
* 安装PaddlePaddle
* 下载PaddleOCR并安装依赖
**2、执行预测**
* 将推理模型下载到jetson
* 执行检测、识别、串联预测即可
详细[参考流程](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.5/deploy/Jetson/readme_ch.md)。
# 9. 总结
检测实验分别使用PP-OCRv3预训练模型在PCB数据集上进行了直接评估、验证集padding、 fine-tune 3种方案，识别实验分别使用PP-OCRv3预训练模型在PCB数据集上进行了直接评估、 fine-tune、添加公开通用识别数据集、增加PCB图片数量4种方案，指标对比如下：
* 检测
| 序号 | 方案 | hmean | 效果提升 | 实验分析 |
| ---- | -------------------------------------------------------- | ------ | -------- | ------------------------------------- |
| 1 | PP-OCRv3英文超轻量检测预训练模型直接评估 | 64.64% | - | 提供的预训练模型具有泛化能力 |
| 2 | PP-OCRv3英文超轻量检测预训练模型 + 验证集padding直接评估 | 72.13% | +7.5% | padding可以提升尺寸较小图片的检测效果 |
| 3 | PP-OCRv3英文超轻量检测预训练模型 + fine-tune | 100% | +27.9% | fine-tune会提升垂类场景效果 |
* 识别
| 序号 | 方案 | acc | 效果提升 | 实验分析 |
| ---- | ------------------------------------------------------------ | ------ | -------- | ------------------------------------------------------------ |
| 1 | PP-OCRv3中英文超轻量识别预训练模型直接评估 | 46.67% | - | 提供的预训练模型具有泛化能力 |
| 2 | PP-OCRv3中英文超轻量识别预训练模型 + fine-tune | 42.02% | -4.6% | 在数据量不足的情况，反而比预训练模型效果低(也可以通过调整超参数再试试) |
| 3 | PP-OCRv3中英文超轻量识别预训练模型 + fine-tune + 公开通用识别数据集 | 77% | +30% | 在数据量不足的情况下，可以考虑补充公开数据训练 |
| 4 | PP-OCRv3中英文超轻量识别预训练模型 + fine-tune + 增加PCB图像数量 | 99.99% | +23% | 如果能获取更多数据量的情况，可以通过增加数据量提升效果 |
* 端到端
| det | rec | fmeasure |
| --------------------------------------------- | ------------------------------------------------------------ | -------- |
| PP-OCRv3英文超轻量检测预训练模型 + fine-tune | PP-OCRv3中英文超轻量识别预训练模型 + fine-tune + 增加PCB图像数量 | 93.3% |
*结论*
PP-OCRv3的检测模型在未经过fine-tune的情况下，在PCB数据集上也有64.64%的精度，说明具有泛化能力。验证集padding之后，精度提升7.5%，在图片尺寸较小的情况，我们可以通过padding的方式提升检测效果。经过 fine-tune 后能够极大的提升检测效果，精度达到100%。
PP-OCRv3的识别模型方案1和方案2对比可以发现，当数据量不足的情况，预训练模型精度可能比fine-tune效果还要高，所以我们可以先尝试预训练模型直接评估。如果在数据量不足的情况下想进一步提升模型效果，可以通过添加公开通用识别数据集，识别效果提升30%，非常有效。最后如果我们能够采集足够多的真实场景数据集，可以通过增加数据量提升模型效果，精度达到99.99%。
# 更多资源
- 更多深度学习知识、产业案例、面试宝典等，请参考：[awesome-DeepLearning](https://github.com/paddlepaddle/awesome-DeepLearning)
- 更多PaddleOCR使用教程，请参考：[PaddleOCR](https://github.com/PaddlePaddle/PaddleOCR/tree/dygraph)
- 飞桨框架相关资料，请参考：[飞桨深度学习平台](https://www.paddlepaddle.org.cn/?fr=paddleEdu_aistudio)
# 参考
* 数据生成代码库：https://github.com/zcswdt/Color_OCR_image_generator
applications/PCB字符识别/gen_data/corpus/text.txt 0 → 100644
浏览文件 @ 5016894e
5ZQ
I4UL
PWL
SNOG
ZL02
1C30
O3H
YHRS
N03S
1U5Y
JTK
EN4F
YKJ
DWNH
R42W
X0V
4OF5
08AM
Y93S
GWE2
0KR
9U2A
DBQ
Y6J
ROZ
K06
KIEY
NZQJ
UN1B
6X4
\ No newline at end of file
applications/PCB字符识别/gen_data/gen.py 0 → 100644
浏览文件 @ 5016894e
# 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/zcswdt/Color_OCR_image_generator
"""
import os
import random
from PIL import Image, ImageDraw, ImageFont
import json
import argparse
def get_char_lines(txt_root_path):
"""
desc:get corpus line
"""
txt_files = os.listdir(txt_root_path)
char_lines = []
for txt in txt_files:
f = open(os.path.join(txt_root_path, txt), mode='r', encoding='utf-8')
lines = f.readlines()
f.close()
for line in lines:
char_lines.append(line.strip())
return char_lines
def get_horizontal_text_picture(image_file, chars, fonts_list, cf):
"""
desc:gen horizontal text picture
"""
img = Image.open(image_file)
if img.mode != 'RGB':
img = img.convert('RGB')
img_w, img_h = img.size
# random choice font
font_path = random.choice(fonts_list)
# random choice font size
font_size = random.randint(cf.font_min_size, cf.font_max_size)
font = ImageFont.truetype(font_path, font_size)
ch_w = []
ch_h = []
for ch in chars:
wt, ht = font.getsize(ch)
ch_w.append(wt)
ch_h.append(ht)
f_w = sum(ch_w)
f_h = max(ch_h)
# add space
char_space_width = max(ch_w)
f_w += (char_space_width * (len(chars) - 1))
x1 = random.randint(0, img_w - f_w)
y1 = random.randint(0, img_h - f_h)
x2 = x1 + f_w
y2 = y1 + f_h
crop_y1 = y1
crop_x1 = x1
crop_y2 = y2
crop_x2 = x2
best_color = (0, 0, 0)
draw = ImageDraw.Draw(img)
for i, ch in enumerate(chars):
draw.text((x1, y1), ch, best_color, font=font)
x1 += (ch_w[i] + char_space_width)
crop_img = img.crop((crop_x1, crop_y1, crop_x2, crop_y2))
return crop_img, chars
def get_vertical_text_picture(image_file, chars, fonts_list, cf):
"""
desc:gen vertical text picture
"""
img = Image.open(image_file)
if img.mode != 'RGB':
img = img.convert('RGB')
img_w, img_h = img.size
# random choice font
font_path = random.choice(fonts_list)
# random choice font size
font_size = random.randint(cf.font_min_size, cf.font_max_size)
font = ImageFont.truetype(font_path, font_size)
ch_w = []
ch_h = []
for ch in chars:
wt, ht = font.getsize(ch)
ch_w.append(wt)
ch_h.append(ht)
f_w = max(ch_w)
f_h = sum(ch_h)
x1 = random.randint(0, img_w - f_w)
y1 = random.randint(0, img_h - f_h)
x2 = x1 + f_w
y2 = y1 + f_h
crop_y1 = y1
crop_x1 = x1
crop_y2 = y2
crop_x2 = x2
best_color = (0, 0, 0)
draw = ImageDraw.Draw(img)
i = 0
for ch in chars:
draw.text((x1, y1), ch, best_color, font=font)
y1 = y1 + ch_h[i]
i = i + 1
crop_img = img.crop((crop_x1, crop_y1, crop_x2, crop_y2))
crop_img = crop_img.transpose(Image.ROTATE_90)
return crop_img, chars
def get_fonts(fonts_path):
"""
desc: get all fonts
"""
font_files = os.listdir(fonts_path)
fonts_list=[]
for font_file in font_files:
font_path=os.path.join(fonts_path, font_file)
fonts_list.append(font_path)
return fonts_list
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--num_img', type=int, default=30, help="Number of images to generate")
parser.add_argument('--font_min_size', type=int, default=11)
parser.add_argument('--font_max_size', type=int, default=12,
help="Help adjust the size of the generated text and the size of the picture")
parser.add_argument('--bg_path', type=str, default='./background',
help='The generated text pictures will be pasted onto the pictures of this folder')
parser.add_argument('--det_bg_path', type=str, default='./det_background',
help='The generated text pictures will use the pictures of this folder as the background')
parser.add_argument('--fonts_path', type=str, default='../../StyleText/fonts',
help='The font used to generate the picture')
parser.add_argument('--corpus_path', type=str, default='./corpus',
help='The corpus used to generate the text picture')
parser.add_argument('--output_dir', type=str, default='./output/', help='Images save dir')
cf = parser.parse_args()
# save path
if not os.path.exists(cf.output_dir):
os.mkdir(cf.output_dir)
# get corpus
txt_root_path = cf.corpus_path
char_lines = get_char_lines(txt_root_path=txt_root_path)
# get all fonts
fonts_path = cf.fonts_path
fonts_list = get_fonts(fonts_path)
# rec bg
img_root_path = cf.bg_path
imnames=os.listdir(img_root_path)
# det bg
det_bg_path = cf.det_bg_path
bg_pics = os.listdir(det_bg_path)
# OCR det files
det_val_file = open(cf.output_dir + 'det_gt_val.txt', 'w', encoding='utf-8')
det_train_file = open(cf.output_dir + 'det_gt_train.txt', 'w', encoding='utf-8')
# det imgs
det_save_dir = 'imgs/'
if not os.path.exists(cf.output_dir + det_save_dir):
os.mkdir(cf.output_dir + det_save_dir)
det_val_save_dir = 'imgs_val/'
if not os.path.exists(cf.output_dir + det_val_save_dir):
os.mkdir(cf.output_dir + det_val_save_dir)
# OCR rec files
rec_val_file = open(cf.output_dir + 'rec_gt_val.txt', 'w', encoding='utf-8')
rec_train_file = open(cf.output_dir + 'rec_gt_train.txt', 'w', encoding='utf-8')
# rec imgs
rec_save_dir = 'rec_imgs/'
if not os.path.exists(cf.output_dir + rec_save_dir):
os.mkdir(cf.output_dir + rec_save_dir)
rec_val_save_dir = 'rec_imgs_val/'
if not os.path.exists(cf.output_dir + rec_val_save_dir):
os.mkdir(cf.output_dir + rec_val_save_dir)
val_ratio = cf.num_img * 0.2 # val dataset ratio
print('start generating...')
for i in range(0, cf.num_img):
imname = random.choice(imnames)
img_path = os.path.join(img_root_path, imname)
rnd = random.random()
# gen horizontal text picture
if rnd < 0.5:
gen_img, chars = get_horizontal_text_picture(img_path, char_lines[i], fonts_list, cf)
ori_w, ori_h = gen_img.size
gen_img = gen_img.crop((0, 3, ori_w, ori_h))
# gen vertical text picture
else:
gen_img, chars = get_vertical_text_picture(img_path, char_lines[i], fonts_list, cf)
ori_w, ori_h = gen_img.size
gen_img = gen_img.crop((3, 0, ori_w, ori_h))
ori_w, ori_h = gen_img.size
# rec imgs
save_img_name = str(i).zfill(4) + '.jpg'
if i < val_ratio:
save_dir = os.path.join(rec_val_save_dir, save_img_name)
line = save_dir + '\t' + char_lines[i] + '\n'
rec_val_file.write(line)
else:
save_dir = os.path.join(rec_save_dir, save_img_name)
line = save_dir + '\t' + char_lines[i] + '\n'
rec_train_file.write(line)
gen_img.save(cf.output_dir + save_dir, quality = 95, subsampling=0)
# det img
# random choice bg
bg_pic = random.sample(bg_pics, 1)[0]
det_img = Image.open(os.path.join(det_bg_path, bg_pic))
# the PCB position is fixed, modify it according to your own scenario
if bg_pic == '1.png':
x1 = 38
y1 = 3
else:
x1 = 34
y1 = 1
det_img.paste(gen_img, (x1, y1))
# text pos
chars_pos = [[x1, y1], [x1 + ori_w, y1], [x1 + ori_w, y1 + ori_h], [x1, y1 + ori_h]]
label = [{"transcription":char_lines[i], "points":chars_pos}]
if i < val_ratio:
save_dir = os.path.join(det_val_save_dir, save_img_name)
det_val_file.write(save_dir + '\t' + json.dumps(
label, ensure_ascii=False) + '\n')
else:
save_dir = os.path.join(det_save_dir, save_img_name)
det_train_file.write(save_dir + '\t' + json.dumps(
label, ensure_ascii=False) + '\n')
det_img.save(cf.output_dir + save_dir, quality = 95, subsampling=0)
applications/车牌识别.md 已删除 100644 → 0
浏览文件 @ 16bafede
# 一种基于PaddleOCR的轻量级车牌识别模型
- [1. 项目介绍](#1-项目介绍)
- [2. 环境搭建](#2-环境搭建)
- [3. 数据集准备](#3-数据集准备)
- [3.1 数据集标注规则](#31-数据集标注规则)
- [3.2 制作符合PP-OCR训练格式的标注文件](#32-制作符合pp-ocr训练格式的标注文件)
- [4. 实验](#4-实验)
- [4.1 检测](#41-检测)
- [4.1.1 预训练模型直接预测](#411-预训练模型直接预测)
- [4.1.2 CCPD车牌数据集fine-tune](#412-ccpd车牌数据集fine-tune)
- [4.1.3 CCPD车牌数据集fine-tune+量化训练](#413-ccpd车牌数据集fine-tune量化训练)
- [4.1.4 模型导出](#414-模型导出)
- [4.2 识别](#42-识别)
- [4.2.1 预训练模型直接预测](#421-预训练模型直接预测)
- [4.2.2 预训练模型直接预测+改动后处理](#422-预训练模型直接预测改动后处理)
- [4.2.3 CCPD车牌数据集fine-tune](#423-ccpd车牌数据集fine-tune)
- [4.2.4 CCPD车牌数据集fine-tune+量化训练](#424-ccpd车牌数据集fine-tune量化训练)
- [4.2.5 模型导出](#425-模型导出)
- [4.3 串联推理](#43-串联推理)
- [4.4 实验总结](#44-实验总结)
## 1. 项目介绍
车牌识别(Vehicle License Plate Recognition，VLPR) 是计算机视频图像识别技术在车辆牌照识别中的一种应用。车牌识别技术要求能够将运动中的汽车牌照从复杂背景中提取并识别出来，在高速公路车辆管理，停车场管理和中得到广泛应用。
结合我国国情，目前车牌识别技术的难点有:
1. 车牌样式多。我国车牌颜色大致有四种：黄底黑字、蓝底白字、白底黑字、黑底白字；车牌格式包括民用车牌、武警车牌、军车车牌、外交车牌、特种车牌、消防车牌等等。
2. 车牌位置不固定。由于不同汽车品牌公司出产的汽车型号和外形各有不同，每辆车的车牌悬挂位置也不一样；
3. 图像质量差: 运动模糊,由于强光，反射或阴影造成的光照和对比度较差, 车牌(部分)遮挡；
4. 在车辆管理等场景场景对于模型速度有着一定限制。
针对以上问题， 本例选用 [PP-OCRv3](../doc/doc_ch/PP-OCRv3_introduction.md) 这一开源超轻量OCR系统进行车牌识别系统的开发。基于PP-OCRv3模型，在CCPD数据集达到99%的检测和94%的识别精度，模型大小12.8M(2.5M+10.3M)。基于量化对模型体积进行进一步压缩到5.8M(1M+4.8M), 同时推理速度提升x%。
aistudio项目链接: [基于PaddleOCR的轻量级车牌识别范例](https://aistudio.baidu.com/aistudio/projectdetail/3919091?contributionType=1)
## 2. 环境搭建
本任务基于Aistudio完成, 具体环境如下：
- 操作系统: Linux
- PaddlePaddle: 2.3
- paddleslim: 2.2.2
- PaddleOCR: Release/2.5
下载 PaddleOCR代码
```bash
git clone -b dygraph https://github.com/PaddlePaddle/PaddleOCR
```
安装依赖库
```bash
pip install -r PaddleOCR/requirements.txt
```
## 3. 数据集准备
所使用的数据集为 CCPD2020 新能源车牌数据集，该数据集为
该数据集分布如下：
|数据集类型|数量|
|---|---|
|训练集| 5769|
|验证集| 1001|
|测试集| 5006|
数据集图片示例如下:
![](https://ai-studio-static-online.cdn.bcebos.com/3bce057a8e0c40a0acbd26b2e29e4e2590a31bc412764be7b9e49799c69cb91c)
数据集可以从这里下载 https://aistudio.baidu.com/aistudio/datasetdetail/101595
下载好数据集后对数据集进行解压
```bash
unzip -d /home/aistudio/data /home/aistudio/data/data101595/CCPD2020.zip
```
### 3.1 数据集标注规则
CPPD数据集的图片文件名具有特殊规则，详细可查看：https://github.com/detectRecog/CCPD
具体规则如下：
例如: 025-95_113-154&383_386&473-386&473_177&454_154&383_363&402-0_0_22_27_27_33_16-37-15.jpg
由分隔符'-'分为几个部分:
- 025:车牌区域与整个图片区域的面积比。
- 95_113: 车牌水平倾斜度和垂直倾斜度, 水平95°, 竖直113°
- 154&383_386&473: 车牌边界框坐标:左上(154, 383), 右下(386, 473)
- 386&473_177&454_154&383_363&402: 车牌四个角点坐标, 坐标顺序为[右下，左下，左上，右上]
- 0_0_22_27_27_33_16: 车牌号码,CCPD中的每个图像只有一个LP。每个LP编号由一个汉字、一个字母和五个字母或数字组成。有效的中国车牌由七个字符组成：省（1个字符）、字母（1个字符）、字母+数字（5个字符）。“0_0_22_27_27_33_16”是每个字符的索引。这三个数组的定义如下。每个数组的最后一个字符是字母O，而不是数字0。我们使用O作为“无字符”的标志，因为中国车牌字符中没有O。
```python
provinces = ["皖", "沪", "津", "渝", "冀", "晋", "蒙", "辽", "吉", "黑", "苏", "浙", "京", "闽", "赣", "鲁", "豫", "鄂", "湘", "粤", "桂", "琼", "川", "贵", "云", "藏", "陕", "甘", "青", "宁", "新", "警", "学", "O"]
alphabets = ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'J', 'K', 'L', 'M', 'N', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W',
'X', 'Y', 'Z', 'O']
ads = ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'J', 'K', 'L', 'M', 'N', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X',
'Y', 'Z', '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 'O']
```
### 3.2 制作符合PP-OCR训练格式的标注文件
在开始训练之前，可使用如下代码制作符合PP-OCR训练格式的标注文件。
```python
import cv2
import os
import json
from tqdm import tqdm
import numpy as np
provinces = ["皖", "沪", "津", "渝", "冀", "晋", "蒙", "辽", "吉", "黑", "苏", "浙", "京", "闽", "赣", "鲁", "豫", "鄂", "湘", "粤", "桂", "琼", "川", "贵", "云", "藏", "陕", "甘", "青", "宁", "新", "警", "学", "O"]
alphabets = ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'J', 'K', 'L', 'M', 'N', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', 'O']
ads = ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'J', 'K', 'L', 'M', 'N', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 'O']
def make_label(img_dir, save_gt_folder, phase):
crop_img_save_dir = os.path.join(save_gt_folder, phase, 'crop_imgs')
os.makedirs(crop_img_save_dir, exist_ok=True)
f_det = open(os.path.join(save_gt_folder, phase, 'det.txt'), 'w', encoding='utf-8')
f_rec = open(os.path.join(save_gt_folder, phase, 'rec.txt'), 'w', encoding='utf-8')
i = 0
for filename in tqdm(os.listdir(os.path.join(img_dir, phase))):
str_list = filename.split('-')
if len(str_list) < 5:
continue
coord_list = str_list[3].split('_')
txt_list = str_list[4].split('_')
boxes = []
for coord in coord_list:
boxes.append([int(x) for x in coord.split("&")])
boxes = [boxes[2], boxes[3], boxes[0], boxes[1]]
lp_number = provinces[int(txt_list[0])] + alphabets[int(txt_list[1])] + ''.join([ads[int(x)] for x in txt_list[2:]])
# det
det_info = [{'points':boxes, 'transcription':lp_number}]
f_det.write('{}\t{}\n'.format(os.path.join(phase, filename), json.dumps(det_info, ensure_ascii=False)))
# rec
boxes = np.float32(boxes)
img = cv2.imread(os.path.join(img_dir, phase, filename))
# crop_img = img[int(boxes[:,1].min()):int(boxes[:,1].max()),int(boxes[:,0].min()):int(boxes[:,0].max())]
crop_img = get_rotate_crop_image(img, boxes)
crop_img_save_filename = '{}_{}.jpg'.format(i,'_'.join(txt_list))
crop_img_save_path = os.path.join(crop_img_save_dir, crop_img_save_filename)
cv2.imwrite(crop_img_save_path, crop_img)
f_rec.write('{}/crop_imgs/{}\t{}\n'.format(phase, crop_img_save_filename, lp_number))
i+=1
f_det.close()
f_rec.close()
def get_rotate_crop_image(img, points):
'''
img_height, img_width = img.shape[0:2]
left = int(np.min(points[:, 0]))
right = int(np.max(points[:, 0]))
top = int(np.min(points[:, 1]))
bottom = int(np.max(points[:, 1]))
img_crop = img[top:bottom, left:right, :].copy()
points[:, 0] = points[:, 0] - left
points[:, 1] = points[:, 1] - top
'''
assert len(points) == 4, "shape of points must be 4*2"
img_crop_width = int(
max(
np.linalg.norm(points[0] - points[1]),
np.linalg.norm(points[2] - points[3])))
img_crop_height = int(
max(
np.linalg.norm(points[0] - points[3]),
np.linalg.norm(points[1] - points[2])))
pts_std = np.float32([[0, 0], [img_crop_width, 0],
[img_crop_width, img_crop_height],
[0, img_crop_height]])
M = cv2.getPerspectiveTransform(points, pts_std)
dst_img = cv2.warpPerspective(
img,
M, (img_crop_width, img_crop_height),
borderMode=cv2.BORDER_REPLICATE,
flags=cv2.INTER_CUBIC)
dst_img_height, dst_img_width = dst_img.shape[0:2]
if dst_img_height * 1.0 / dst_img_width >= 1.5:
dst_img = np.rot90(dst_img)
return dst_img
img_dir = '/home/aistudio/data/CCPD2020/ccpd_green'
save_gt_folder = '/home/aistudio/data/CCPD2020/PPOCR'
# phase = 'train' # change to val and test to make val dataset and test dataset
for phase in ['train','val','test']:
make_label(img_dir, save_gt_folder, phase)
```
通过上述命令可以完成了`训练集`，`验证集`和`测试集`的制作，制作完成的数据集信息如下：
| 类型 | 数据集 | 图片地址 | 标签地址 | 图片数量 |
| --- | --- | --- | --- | --- |
| 检测 | 训练集 | /home/aistudio/data/CCPD2020/ccpd_green/train | /home/aistudio/data/CCPD2020/PPOCR/train/det.txt | 5769 |
| 检测 | 验证集 | /home/aistudio/data/CCPD2020/ccpd_green/val | /home/aistudio/data/CCPD2020/PPOCR/val/det.txt | 1001 |
| 检测 | 测试集 | /home/aistudio/data/CCPD2020/ccpd_green/test | /home/aistudio/data/CCPD2020/PPOCR/test/det.txt | 5006 |
| 识别 | 训练集 | /home/aistudio/data/CCPD2020/PPOCR/train/crop_imgs | /home/aistudio/data/CCPD2020/PPOCR/train/rec.txt | 5769 |
| 识别 | 验证集 | /home/aistudio/data/CCPD2020/PPOCR/val/crop_imgs | /home/aistudio/data/CCPD2020/PPOCR/val/rec.txt | 1001 |
| 识别 | 测试集 | /home/aistudio/data/CCPD2020/PPOCR/test/crop_imgs | /home/aistudio/data/CCPD2020/PPOCR/test/rec.txt | 5006 |
在普遍的深度学习流程中，都是在训练集训练，在验证集选择最优模型后在测试集上进行测试。在本例中，我们省略中间步骤，直接在训练集训练，在测试集选择最优模型，因此我们只使用训练集和测试集。
## 4. 实验
由于数据集比较少，为了模型更好和更快的收敛，这里选用 PaddleOCR 中的 PP-OCRv3 模型进行文本检测和识别，并且使用 PP-OCRv3 模型参数作为预训练模型。PP-OCRv3在PP-OCRv2的基础上，中文场景端到端Hmean指标相比于PP-OCRv2提升5%, 英文数字模型端到端效果提升11%。详细优化细节请参考[PP-OCRv3](../doc/doc_ch/PP-OCRv3_introduction.md)技术报告。
由于车牌场景均为端侧设备部署，因此对速度和模型大小有比较高的要求，因此还需要采用量化训练的方式进行模型大小的压缩和模型推理速度的加速。模型量化可以在基本不损失模型的精度的情况下，将FP32精度的模型参数转换为Int8精度，减小模型参数大小并加速计算，使用量化后的模型在移动端等部署时更具备速度优势。
因此，本实验中对于车牌检测和识别有如下3种方案：
1. PP-OCRv3中英文超轻量预训练模型直接预测
2. CCPD车牌数据集在PP-OCRv3模型上fine-tune
3. CCPD车牌数据集在PP-OCRv3模型上fine-tune后量化
### 4.1 检测
#### 4.1.1 预训练模型直接预测
从下表中下载PP-OCRv3文本检测预训练模型
|模型名称|模型简介|配置文件|推理模型大小|下载地址|
| --- | --- | --- | --- | --- |
|ch_PP-OCRv3_det| 【最新】原始超轻量模型，支持中英文、多语种文本检测 |[ch_PP-OCRv3_det_cml.yml](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml)| 3.8M |[推理模型](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_distill_train.tar)|
使用如下命令下载预训练模型
```bash
mkdir models
cd models
wget https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_distill_train.tar
tar -xf ch_PP-OCRv3_det_distill_train.tar
cd /home/aistudio/PaddleOCR
```
预训练模型下载完成后，我们使用[ch_PP-OCRv3_det_student.yml](../configs/chepai/ch_PP-OCRv3_det_student.yml) 配置文件进行后续实验，在开始评估之前需要对配置文件中部分字段进行设置，具体如下：
1. 模型存储和训练相关:
1. Global.pretrained_model: 指向PP-OCRv3文本检测预训练模型地址
2. 数据集相关
1. Eval.dataset.data_dir：指向测试集图片存放目录
2. Eval.dataset.label_file_list：指向测试集标注文件
上述字段均为必须修改的字段，可以通过修改配置文件的方式改动，也可在不需要修改配置文件的情况下，改变训练的参数。这里使用不改变配置文件的方式 。使用如下命令进行PP-OCRv3文本检测预训练模型的评估
```bash
python tools/eval.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_student.yml -o \
Global.pretrained_model=models/ch_PP-OCRv3_det_distill_train/student.pdparams \
Eval.dataset.data_dir=/home/aistudio/data/CCPD2020/ccpd_green \
Eval.dataset.label_file_list=[/home/aistudio/data/CCPD2020/PPOCR/test/det.txt]
```
上述指令中，通过-c 选择训练使用配置文件，通过-o参数在不需要修改配置文件的情况下，改变训练的参数。
使用预训练模型进行评估，指标如下所示：
| 方案 |hmeans|
|---------------------------|---|
| PP-OCRv3中英文超轻量检测预训练模型直接预测 |76.12%|
#### 4.1.2 CCPD车牌数据集fine-tune
**训练**
为了进行fine-tune训练，我们需要在配置文件中设置需要使用的预训练模型地址，学习率和数据集等参数。 具体如下:
1. 模型存储和训练相关:
1. Global.pretrained_model: 指向PP-OCRv3文本检测预训练模型地址
2. Global.eval_batch_step: 模型多少step评估一次，这里设为从第0个step开始没隔772个step评估一次，772为一个epoch总的step数。
2. 优化器相关:
1. Optimizer.lr.name: 学习率衰减器设为常量 Const
2. Optimizer.lr.learning_rate: 做finetune实验，学习率需要设置的比较小，此处学习率设为配置文件中的0.05倍
3. Optimizer.lr.warmup_epoch: warmup_epoch设为0
3. 数据集相关:
1. Train.dataset.data_dir：指向训练集图片存放目录
2. Train.dataset.label_file_list：指向训练集标注文件
3. Eval.dataset.data_dir：指向测试集图片存放目录
4. Eval.dataset.label_file_list：指向测试集标注文件
使用如下代码即可启动在CCPD车牌数据集上的fine-tune。
```bash
python tools/train.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_student.yml -o \
Global.pretrained_model=models/ch_PP-OCRv3_det_distill_train/student.pdparams \
Global.save_model_dir=output/CCPD/det \
Global.eval_batch_step="[0, 772]" \
Optimizer.lr.name=Const \
Optimizer.lr.learning_rate=0.0005 \
Optimizer.lr.warmup_epoch=0 \
Train.dataset.data_dir=/home/aistudio/data/CCPD2020/ccpd_green \
Train.dataset.label_file_list=[/home/aistudio/data/CCPD2020/PPOCR/train/det.txt] \
Eval.dataset.data_dir=/home/aistudio/data/CCPD2020/ccpd_green \
Eval.dataset.label_file_list=[/home/aistudio/data/CCPD2020/PPOCR/test/det.txt]
```
在上述命令中，通过`-o`的方式修改了配置文件中的参数。
训练好的模型地址为: [det_ppocr_v3_finetune.tar](https://paddleocr.bj.bcebos.com/fanliku/license_plate_recognition/det_ppocr_v3_finetune.tar)
**评估**
训练完成后使用如下命令进行评估
```bash
python tools/eval.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_student.yml -o \
Global.pretrained_model=output/CCPD/det/best_accuracy.pdparams \
Eval.dataset.data_dir=/home/aistudio/data/CCPD2020/ccpd_green \
Eval.dataset.label_file_list=[/home/aistudio/data/CCPD2020/PPOCR/test/det.txt]
```
使用预训练模型和CCPD车牌数据集fine-tune，指标分别如下：
|方案|hmeans|
|---|---|
|PP-OCRv3中英文超轻量检测预训练模型直接预测|76.12%|
|PP-OCRv3中英文超轻量检测预训练模型 fine-tune|99%|
可以看到进行fine-tune能显著提升车牌检测的效果。
#### 4.1.3 CCPD车牌数据集fine-tune+量化训练
此处采用 PaddleOCR 中提供好的[量化教程](../deploy/slim/quantization/README.md)对模型进行量化训练。
量化训练可通过如下命令启动:
```bash
python3.7 deploy/slim/quantization/quant.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_student.yml -o \
Global.pretrained_model=output/CCPD/det/best_accuracy.pdparams \
Global.save_model_dir=output/CCPD/det_quant \
Global.eval_batch_step="[0, 772]" \
Optimizer.lr.name=Const \
Optimizer.lr.learning_rate=0.0005 \
Optimizer.lr.warmup_epoch=0 \
Train.dataset.data_dir=/home/aistudio/data/CCPD2020/ccpd_green \
Train.dataset.label_file_list=[/home/aistudio/data/CCPD2020/PPOCR/train/det.txt] \
Eval.dataset.data_dir=/home/aistudio/data/CCPD2020/ccpd_green \
Eval.dataset.label_file_list=[/home/aistudio/data/CCPD2020/PPOCR/test/det.txt]
```
训练好的模型地址为: [det_ppocr_v3_quant.tar](https://paddleocr.bj.bcebos.com/fanliku/license_plate_recognition/det_ppocr_v3_quant.tar)
量化后指标对比如下
|方案|hmeans| 模型大小 |预测速度(lite)|
|---|---|------|---|
|PP-OCRv3中英文超轻量检测预训练模型 fine-tune|99%| 2.5M | 223ms/image |
|PP-OCRv3中英文超轻量检测预训练模型 fine-tune+量化|98.91%| 1M | 189ms/image |
可以看到量化后能显著降低模型体积并且精度几乎无损。
预测速度是在android骁龙855上预测275张图像的平均耗时。模型在移动端的部署步骤参考[文档](../deploy/lite/readme_ch.md)
#### 4.1.4 模型导出
使用如下命令可以将训练好的模型进行导出
* 非量化模型
```bash
python tools/export_model.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_student.yml -o \
Global.pretrained_model=output/CCPD/det/best_accuracy.pdparams \
Global.save_inference_dir=output/det/infer
```
* 量化模型
```bash
python deploy/slim/quantization/export_model.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_student.yml -o \
Global.pretrained_model=output/CCPD/det_quant/best_accuracy.pdparams \
Global.save_inference_dir=output/det/infer
```
### 4.2 识别
#### 4.2.1 预训练模型直接预测
从下表中下载PP-OCRv3文本识别预训练模型
|模型名称|模型简介|配置文件|推理模型大小|下载地址|
| --- | --- | --- | --- | --- |
|ch_PP-OCRv3_rec|【最新】原始超轻量模型，支持中英文、数字识别|[ch_PP-OCRv3_rec_distillation.yml](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml)| 12.4M |[推理模型](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_train.tar) |
使用如下命令下载预训练模型
```bash
mkdir models
cd models
wget https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_train.tar
tar -xf ch_PP-OCRv3_rec_train.tar
cd /home/aistudio/PaddleOCR
```
PaddleOCR提供的PP-OCRv3识别模型采用蒸馏训练策略，因此提供的预训练模型中会包含`Teacher`和`Student`模型的参数，详细信息可参考[knowledge_distillation.md](../doc/doc_ch/knowledge_distillation.md)。 因此，模型下载完成后需要使用如下代码提取`Student`模型的参数：
```python
import paddle
# 加载预训练模型
all_params = paddle.load("models/ch_PP-OCRv3_rec_train/best_accuracy.pdparams")
# 查看权重参数的keys
print(all_params.keys())
# 学生模型的权重提取
s_params = {key[len("Student."):]: all_params[key] for key in all_params if "Student." in key}
# 查看学生模型权重参数的keys
print(s_params.keys())
# 保存
paddle.save(s_params, "models/ch_PP-OCRv3_rec_train/student.pdparams")
```
预训练模型下载完成后，我们使用[ch_PP-OCRv3_rec.yml](../configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml) 配置文件进行后续实验，在开始评估之前需要对配置文件中部分字段进行设置，具体如下：
1. 模型存储和训练相关:
1. Global.pretrained_model: 指向PP-OCRv3文本识别预训练模型地址
2. 数据集相关
1. Eval.dataset.data_dir：指向测试集图片存放目录
2. Eval.dataset.label_file_list：指向测试集标注文件
使用如下命令进行PP-OCRv3文本识别预训练模型的评估
```bash
python tools/eval.py -c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml -o \
Global.pretrained_model=models/ch_PP-OCRv3_rec_train/student.pdparams \
Eval.dataset.data_dir=/home/aistudio/data/CCPD2020/PPOCR \
Eval.dataset.label_file_list=[/home/aistudio/data/CCPD2020/PPOCR/test/rec.txt]
```
评估部分日志如下：
```bash
[2022/05/12 19:52:02] ppocr INFO: load pretrain successful from models/ch_PP-OCRv3_rec_train/best_accuracy
eval model:: 100%|██████████████████████████████| 40/40 [00:15<00:00, 2.57it/s]
[2022/05/12 19:52:17] ppocr INFO: metric eval ***************
[2022/05/12 19:52:17] ppocr INFO: acc:0.0
[2022/05/12 19:52:17] ppocr INFO: norm_edit_dis:0.8656084923002452
[2022/05/12 19:52:17] ppocr INFO: Teacher_acc:0.000399520574511545
[2022/05/12 19:52:17] ppocr INFO: Teacher_norm_edit_dis:0.8657902943394548
[2022/05/12 19:52:17] ppocr INFO: fps:1443.1801978719905
```
使用预训练模型进行评估，指标如下所示：
|方案|acc|
|---|---|
|PP-OCRv3中英文超轻量识别预训练模型直接预测|0%|
从评估日志中可以看到，直接使用PP-OCRv3预训练模型进行评估，acc非常低，但是norm_edit_dis很高。因此，我们猜测是模型大部分文字识别是对的，只有少部分文字识别错误。使用如下命令进行infer查看模型的推理结果进行验证：
```bash
python tools/infer_rec.py -c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml -o \
Global.pretrained_model=models/ch_PP-OCRv3_rec_train/student.pdparams \
Global.infer_img=/home/aistudio/data/CCPD2020/PPOCR/test/crop_imgs/0_0_0_3_32_30_31_30_30.jpg
```
输出部分日志如下：
```bash
[2022/05/01 08:51:57] ppocr INFO: train with paddle 2.2.2 and device CUDAPlace(0)
W0501 08:51:57.127391 11326 device_context.cc:447] Please NOTE: device: 0, GPU Compute Capability: 7.0, Driver API Version: 11.0, Runtime API Version: 10.1
W0501 08:51:57.132315 11326 device_context.cc:465] device: 0, cuDNN Version: 7.6.
[2022/05/01 08:52:00] ppocr INFO: load pretrain successful from models/ch_PP-OCRv3_rec_train/student
[2022/05/01 08:52:00] ppocr INFO: infer_img: /home/aistudio/data/CCPD2020/PPOCR/test/crop_imgs/0_0_3_32_30_31_30_30.jpg
[2022/05/01 08:52:00] ppocr INFO: result: {"Student": {"label": "皖A·D86766", "score": 0.9552637934684753}, "Teacher": {"label": "皖A·D86766", "score": 0.9917094707489014}}
[2022/05/01 08:52:00] ppocr INFO: success!
```
从infer结果可以看到，车牌中的文字大部分都识别正确，只是多识别出了一个`·`。针对这种情况，有如下两种方案：
1. 直接通过后处理去掉多识别的`·`。
2. 进行finetune。
#### 4.2.2 预训练模型直接预测+改动后处理
直接通过后处理去掉多识别的`·`，在后处理的改动比较简单，只需在 [ppocr/postprocess/rec_postprocess.py](../ppocr/postprocess/rec_postprocess.py) 文件的76行添加如下代码:
```python
text = text.replace('·','')
```
改动前后指标对比:
|方案|acc|
|---|---|
|PP-OCRv3中英文超轻量识别预训练模型直接预测|0.2%|
|PP-OCRv3中英文超轻量识别预训练模型直接预测+后处理去掉多识别的`·`|90.97%|
可以看到，去掉多余的`·`能大幅提高精度。
#### 4.2.3 CCPD车牌数据集fine-tune
**训练**
为了进行fine-tune训练，我们需要在配置文件中设置需要使用的预训练模型地址，学习率和数据集等参数。 具体如下:
1. 模型存储和训练相关:
1. Global.pretrained_model: 指向PP-OCRv3文本识别预训练模型地址
2. Global.eval_batch_step: 模型多少step评估一次，这里设为从第0个step开始没隔45个step评估一次，45为一个epoch总的step数。
2. 优化器相关
1. Optimizer.lr.name: 学习率衰减器设为常量 Const
2. Optimizer.lr.learning_rate: 做finetune实验，学习率需要设置的比较小，此处学习率设为配置文件中的0.05倍
3. Optimizer.lr.warmup_epoch: warmup_epoch设为0
3. 数据集相关
1. Train.dataset.data_dir：指向训练集图片存放目录
2. Train.dataset.label_file_list：指向训练集标注文件
3. Eval.dataset.data_dir：指向测试集图片存放目录
4. Eval.dataset.label_file_list：指向测试集标注文件
使用如下命令启动finetune
```bash
python tools/train.py -c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml -o \
Global.pretrained_model=models/ch_PP-OCRv3_rec_train/student.pdparams \
Global.save_model_dir=output/CCPD/rec/ \
Global.eval_batch_step="[0, 90]" \
Optimizer.lr.name=Const \
Optimizer.lr.learning_rate=0.0005 \
Optimizer.lr.warmup_epoch=0 \
Train.dataset.data_dir=/home/aistudio/data/CCPD2020/PPOCR \
Train.dataset.label_file_list=[/home/aistudio/data/CCPD2020/PPOCR/train/rec.txt] \
Eval.dataset.data_dir=/home/aistudio/data/CCPD2020/PPOCR \
Eval.dataset.label_file_list=[/home/aistudio/data/CCPD2020/PPOCR/test/rec.txt]
```
训练好的模型地址为: [rec_ppocr_v3_finetune.tar](https://paddleocr.bj.bcebos.com/fanliku/license_plate_recognition/rec_ppocr_v3_finetune.tar)
**评估**
训练完成后使用如下命令进行评估
```bash
python tools/eval.py -c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml -o \
Global.pretrained_model=output/CCPD/rec/best_accuracy.pdparams \
Eval.dataset.data_dir=/home/aistudio/data/CCPD2020/PPOCR \
Eval.dataset.label_file_list=[/home/aistudio/data/CCPD2020/PPOCR/test/rec.txt]
```
使用预训练模型和CCPD车牌数据集fine-tune，指标分别如下：
|方案| acc |
|---|--------|
|PP-OCRv3中英文超轻量识别预训练模型直接预测| 0% |
|PP-OCRv3中英文超轻量识别预训练模型直接预测+后处理去掉多识别的`·`| 90.97% |
|PP-OCRv3中英文超轻量识别预训练模型 fine-tune| 94.54% |
可以看到进行fine-tune能显著提升车牌识别的效果。
#### 4.2.4 CCPD车牌数据集fine-tune+量化训练
此处采用 PaddleOCR 中提供好的[量化教程](../deploy/slim/quantization/README.md)对模型进行量化训练。
量化训练可通过如下命令启动:
```bash
python3.7 deploy/slim/quantization/quant.py -c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml -o \
Global.pretrained_model=output/CCPD/rec/best_accuracy.pdparams \
Global.save_model_dir=output/CCPD/rec_quant/ \
Global.eval_batch_step="[0, 90]" \
Optimizer.lr.name=Const \
Optimizer.lr.learning_rate=0.0005 \
Optimizer.lr.warmup_epoch=0 \
Train.dataset.data_dir=/home/aistudio/data/CCPD2020/PPOCR \
Train.dataset.label_file_list=[/home/aistudio/data/CCPD2020/PPOCR/train/rec.txt] \
Eval.dataset.data_dir=/home/aistudio/data/CCPD2020/PPOCR \
Eval.dataset.label_file_list=[/home/aistudio/data/CCPD2020/PPOCR/test/rec.txt]
```
训练好的模型地址为: [rec_ppocr_v3_quant.tar](https://paddleocr.bj.bcebos.com/fanliku/license_plate_recognition/rec_ppocr_v3_quant.tar)
量化后指标对比如下
|方案| acc | 模型大小 |预测速度(lite)|
|---|--------|-------|---|
|PP-OCRv3中英文超轻量识别预训练模型 fine-tune| 94.54% | 10.3M | 4.2ms/image |
|PP-OCRv3中英文超轻量识别预训练模型 fine-tune + 量化| 93.4% | 4.8M | 1.8ms/image; |
可以看到量化后能显著降低模型体积，但是由于识别数据过少，量化带来了1%的精度下降。
预测速度是在android骁龙855上预测5006张识别文字图像的平均耗时。模型在移动端的部署步骤参考[文档](../deploy/lite/readme_ch.md)
#### 4.2.5 模型导出
使用如下命令可以将训练好的模型进行导出。
* 非量化模型
```bash
python tools/export_model.py -c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml -o \
Global.pretrained_model=output/CCPD/rec/best_accuracy.pdparams \
Global.save_inference_dir=output/CCPD/rec/infer
```
* 量化模型
```bash
python deploy/slim/quantization/export_model.py -c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml -o \
Global.pretrained_model=output/CCPD/rec_quant/best_accuracy.pdparams \
Global.save_inference_dir=output/CCPD/rec_quant/infer
```
### 4.3 串联推理
检测模型和识别模型分别fine-tune并导出为inference模型之后，可以使用如下命令进行端到端推理并对结果进行可视化。
```bash
python tools/infer/predict_system.py \
--det_model_dir=output/CCPD/det/infer/ \
--rec_model_dir=output/CCPD/rec/infer/ \
--image_dir="/home/aistudio/data/CCPD2020/ccpd_green/test/04131106321839081-92_258-159&509_530&611-527&611_172&599_159&509_530&525-0_0_3_32_30_31_30_30-109-106.jpg" \
--rec_image_shape=3,48,320
```
推理结果如下
![](https://ai-studio-static-online.cdn.bcebos.com/76b6a0939c2c4cf49039b6563c4b28e241e11285d7464e799e81c58c0f7707a7)
### 4.4 实验总结
我们分别使用PP-OCRv3中英文超轻量预训练模型在车牌数据集上进行了直接评估和finetune 和finetune+量化3种方案的实验，指标对比如下：
- 检测
|方案|hmeans| 模型大小 |预测速度(lite)|
|---|---|------|---|
|PP-OCRv3中英文超轻量检测预训练模型直接预测|76.12%|2.5M| 223ms/image |
|PP-OCRv3中英文超轻量检测预训练模型 fine-tune|99%| 2.5M | 223ms/image |
|PP-OCRv3中英文超轻量检测预训练模型 fine-tune+量化|98.91%| 1M | 189ms/image |
预测速度是在android骁龙855上预测275张图像的平均耗时。
- 识别
|方案| acc | 模型大小 |预测速度(lite)|
|---|--------|-------|---|
|PP-OCRv3中英文超轻量识别预训练模型直接预测| 0% |10.3M| 4.2ms/image |
|PP-OCRv3中英文超轻量识别预训练模型直接预测+后处理去掉多识别的`·`| 90.97% |10.3M| 4.2ms/image |
|PP-OCRv3中英文超轻量识别预训练模型 fine-tune| 94.54% | 10.3M | 4.2ms/image |
|PP-OCRv3中英文超轻量识别预训练模型 fine-tune + 量化| 94.4% | 4.8M | 1.8ms/image |
预测速度是在android骁龙855上预测5006张识别文字图像的平均耗时。
- 结论
PP-OCRv3的检测模型在未经过fine-tune的情况下，在车牌数据集上也有一定的精度，经过finetune后能够极大的提升检测效果，精度达到99%。在使用量化训练后检测模型的精度几乎无损，并且模型大小压缩60%。
PP-OCRv3的识别模型在未经过fine-tune的情况下，在车牌数据集上精度为0，但是经过分析可以知道，模型大部分字符都预测正确，但是会多预测一个特殊字符，去掉这个特殊字符后，精度达到90%。PP-OCRv3识别模型在经过finetune后识别精度进一步提升，达到94.4%。在使用量化训练后识别模型大小压缩53%，但是由于数据量多少，带来了1%的精度损失。
applications/轻量级车牌识别.md 0 → 100644
浏览文件 @ 5016894e
因为 它太大了无法显示 source diff 。你可以改为 查看blob。
configs/cls/ch_PP-OCRv3/ch_PP-OCRv3_rotnet.yml
浏览文件 @ 5016894e
......@@ -63,8 +63,7 @@ Train:
- DecodeImage:
img_mode: BGR
channel_first: false
- RecAug:
use_tia: False
- BaseDataAugmentation:
- RandAugment:
- SSLRotateResize:
image_shape: [3, 48, 320]
......
configs/cls/cls_mv3.yml
浏览文件 @ 5016894e
......@@ -60,8 +60,7 @@ Train:
img_mode: BGR
channel_first: False
- ClsLabelEncode: # Class handling label
- RecAug:
use_tia: False
- BaseDataAugmentation:
- RandAugment:
- ClsResizeImg:
image_shape: [3, 48, 192]
......
deploy/cpp_infer/CMakeLists.txt
浏览文件 @ 5016894e
......@@ -92,6 +92,8 @@ include_directories("${PADDLE_LIB}/third_party/install/glog/include")
include_directories("${PADDLE_LIB}/third_party/install/gflags/include")
include_directories("${PADDLE_LIB}/third_party/install/xxhash/include")
include_directories("${PADDLE_LIB}/third_party/install/zlib/include")
include_directories("${PADDLE_LIB}/third_party/install/onnxruntime/include")
include_directories("${PADDLE_LIB}/third_party/install/paddle2onnx/include")
include_directories("${PADDLE_LIB}/third_party/boost")
include_directories("${PADDLE_LIB}/third_party/eigen3")
......@@ -110,6 +112,8 @@ link_directories("${PADDLE_LIB}/third_party/install/protobuf/lib")
link_directories("${PADDLE_LIB}/third_party/install/glog/lib")
link_directories("${PADDLE_LIB}/third_party/install/gflags/lib")
link_directories("${PADDLE_LIB}/third_party/install/xxhash/lib")
link_directories("${PADDLE_LIB}/third_party/install/onnxruntime/lib")
link_directories("${PADDLE_LIB}/third_party/install/paddle2onnx/lib")
link_directories("${PADDLE_LIB}/paddle/lib")
......
deploy/cpp_infer/readme.md
浏览文件 @ 5016894e
......@@ -208,7 +208,7 @@ Execute the built executable file:
./build/ppocr [--param1] [--param2] [...]
```
**Note**:ppocr uses the `PP-OCRv3` model by default, and the input shape used by the recognition model is `3, 48, 320`, so if you use the recognition function, you need to add the parameter `--rec_img_h=48`, if you do not use the default `PP-OCRv3` model, you do not need to set this parameter.
**Note**:ppocr uses the `PP-OCRv3` model by default, and the input shape used by the recognition model is `3, 48, 320`, if you want to use the old version model, you should add the parameter `--rec_img_h=32`.
Specifically,
......@@ -222,7 +222,6 @@ Specifically,
--det=true \
--rec=true \
--cls=true \
--rec_img_h=48\
```
##### 2. det+rec：
......@@ -234,7 +233,6 @@ Specifically,
--det=true \
--rec=true \
--cls=false \
--rec_img_h=48\
```
##### 3. det
......@@ -254,7 +252,6 @@ Specifically,
--det=false \
--rec=true \
--cls=true \
--rec_img_h=48\
```
##### 5. rec
......@@ -265,7 +262,6 @@ Specifically,
--det=false \
--rec=true \
--cls=false \
--rec_img_h=48\
```
##### 6. cls
......@@ -330,7 +326,7 @@ More parameters are as follows,
|rec_model_dir|string|-|Address of recognition inference model|
|rec_char_dict_path|string|../../ppocr/utils/ppocr_keys_v1.txt|dictionary file|
|rec_batch_num|int|6|batch size of recognition|
|rec_img_h|int|32|image height of recognition|
|rec_img_h|int|48|image height of recognition|
|rec_img_w|int|320|image width of recognition|
* Multi-language inference is also supported in PaddleOCR, you can refer to [recognition tutorial](../../doc/doc_en/recognition_en.md) for more supported languages and models in PaddleOCR. Specifically, if you want to infer using multi-language models, you just need to modify values of `rec_char_dict_path` and `rec_model_dir`.
......
deploy/cpp_infer/readme_ch.md
浏览文件 @ 5016894e
......@@ -213,7 +213,7 @@ CUDNN_LIB_DIR=/your_cudnn_lib_dir
本demo支持系统串联调用，也支持单个功能的调用，如，只使用检测或识别功能。
**注意** ppocr默认使用`PP-OCRv3`模型，识别模型使用的输入shape为`3,48,320`, 因此如果使用识别功能，需要添加参数`--rec_img_h=48`，如果不使用默认的`PP-OCRv3`模型，则无需设置该参数。
**注意** ppocr默认使用`PP-OCRv3`模型，识别模型使用的输入shape为`3,48,320`, 如需使用旧版本的PP-OCR模型，则需要设置参数`--rec_img_h=32`。
运行方式：
......@@ -232,7 +232,6 @@ CUDNN_LIB_DIR=/your_cudnn_lib_dir
--det=true \
--rec=true \
--cls=true \
--rec_img_h=48\
```
##### 2. 检测+识别：
......@@ -244,7 +243,6 @@ CUDNN_LIB_DIR=/your_cudnn_lib_dir
--det=true \
--rec=true \
--cls=false \
--rec_img_h=48\
```
##### 3. 检测：
......@@ -264,7 +262,6 @@ CUDNN_LIB_DIR=/your_cudnn_lib_dir
--det=false \
--rec=true \
--cls=true \
--rec_img_h=48\
```
##### 5. 识别：
......@@ -275,7 +272,6 @@ CUDNN_LIB_DIR=/your_cudnn_lib_dir
--det=false \
--rec=true \
--cls=false \
--rec_img_h=48\
```
##### 6. 分类：
......@@ -339,7 +335,7 @@ CUDNN_LIB_DIR=/your_cudnn_lib_dir
|rec_model_dir|string|-|识别模型inference model地址|
|rec_char_dict_path|string|../../ppocr/utils/ppocr_keys_v1.txt|字典文件|
|rec_batch_num|int|6|识别模型batchsize|
|rec_img_h|int|32|识别模型输入图像高度|
|rec_img_h|int|48|识别模型输入图像高度|
|rec_img_w|int|320|识别模型输入图像宽度|
......
deploy/cpp_infer/src/args.cpp
浏览文件 @ 5016894e
......@@ -47,7 +47,7 @@ DEFINE_string(rec_model_dir, "", "Path of rec inference model.");
DEFINE_int32(rec_batch_num, 6, "rec_batch_num.");
DEFINE_string(rec_char_dict_path, "../../ppocr/utils/ppocr_keys_v1.txt",
"Path of dictionary.");
DEFINE_int32(rec_img_h, 32, "rec image height");
DEFINE_int32(rec_img_h, 48, "rec image height");
DEFINE_int32(rec_img_w, 320, "rec image width");
// ocr forward related
......
deploy/cpp_infer/src/ocr_rec.cpp
浏览文件 @ 5016894e
......@@ -83,7 +83,7 @@ void CRNNRecognizer::Run(std::vector<cv::Mat> img_list,
int out_num = std::accumulate(predict_shape.begin(), predict_shape.end(), 1,
std::multiplies<int>());
predict_batch.resize(out_num);
// predict_batch is the result of Last FC with softmax
output_t->CopyToCpu(predict_batch.data());
auto inference_end = std::chrono::steady_clock::now();
inference_diff += inference_end - inference_start;
......@@ -98,9 +98,11 @@ void CRNNRecognizer::Run(std::vector<cv::Mat> img_list,
float max_value = 0.0f;
for (int n = 0; n < predict_shape[1]; n++) {
// get idx
argmax_idx = int(Utility::argmax(
&predict_batch[(m * predict_shape[1] + n) * predict_shape[2]],
&predict_batch[(m * predict_shape[1] + n + 1) * predict_shape[2]]));
// get score
max_value = float(*std::max_element(
&predict_batch[(m * predict_shape[1] + n) * predict_shape[2]],
&predict_batch[(m * predict_shape[1] + n + 1) * predict_shape[2]]));
......@@ -132,7 +134,9 @@ void CRNNRecognizer::LoadModel(const std::string &model_dir) {
paddle_infer::Config config;
config.SetModel(model_dir + "/inference.pdmodel",
model_dir + "/inference.pdiparams");
std::cout << "In PP-OCRv3, default rec_img_h is 48,"
<< "if you use other model, you should set the param rec_img_h=32"
<< std::endl;
if (this->use_gpu_) {
config.EnableUseGpu(this->gpu_mem_, this->gpu_id_);
if (this->use_tensorrt_) {
......
deploy/pdserving/README.md
浏览文件 @ 5016894e
......@@ -136,7 +136,7 @@ The recognition model is the same.
2. Run the following command to start the service.
```
# Start the service and save the running log in log.txt
python3 web_service.py &>log.txt &
python3 web_service.py --config=config.yml &>log.txt &
```
After the service is successfully started, a log similar to the following will be printed in log.txt
![](./imgs/start_server.png)
......@@ -217,7 +217,7 @@ The C++ service deployment is the same as python in the environment setup and da
2. Run the following command to start the service.
```
# Start the service and save the running log in log.txt
python3 -m paddle_serving_server.serve --model ppocr_det_v3_serving ppocr_rec_v3_serving --op GeneralDetectionOp GeneralInferOp --port 9293 &>log.txt &
python3 -m paddle_serving_server.serve --model ppocr_det_v3_serving ppocr_rec_v3_serving --op GeneralDetectionOp GeneralInferOp --port 8181 &>log.txt &
```
After the service is successfully started, a log similar to the following will be printed in log.txt
![](./imgs/start_server.png)
......
deploy/pdserving/README_CN.md
浏览文件 @ 5016894e
......@@ -135,7 +135,7 @@ python3 -m paddle_serving_client.convert --dirname ./ch_PP-OCRv3_rec_infer/ \
2. 启动服务可运行如下命令：
```
# 启动服务，运行日志保存在log.txt
python3 web_service.py &>log.txt &
python3 web_service.py --config=config.yml &>log.txt &
```
成功启动服务后，log.txt中会打印类似如下日志
![](./imgs/start_server.png)
......@@ -230,7 +230,7 @@ cp -rf general_detection_op.cpp Serving/core/general-server/op
```
# 启动服务，运行日志保存在log.txt
python3 -m paddle_serving_server.serve --model ppocr_det_v3_serving ppocr_rec_v3_serving --op GeneralDetectionOp GeneralInferOp --port 9293 &>log.txt &
python3 -m paddle_serving_server.serve --model ppocr_det_v3_serving ppocr_rec_v3_serving --op GeneralDetectionOp GeneralInferOp --port 8181 &>log.txt &
```
成功启动服务后，log.txt中会打印类似如下日志
![](./imgs/start_server.png)
......
deploy/pdserving/ocr_cpp_client.py
浏览文件 @ 5016894e
......@@ -22,15 +22,16 @@ import cv2
from paddle_serving_app.reader import Sequential, URL2Image, ResizeByFactor
from paddle_serving_app.reader import Div, Normalize, Transpose
from ocr_reader import OCRReader
import codecs
client = Client()
# TODO:load_client need to load more than one client model.
# this need to figure out some details.
client.load_client_config(sys.argv[1:])
client.connect(["127.0.0.1:9293"])
client.connect(["127.0.0.1:8181"])
import paddle
test_img_dir = "../../doc/imgs/"
test_img_dir = "../../doc/imgs/1.jpg"
ocr_reader = OCRReader(char_dict_path="../../ppocr/utils/ppocr_keys_v1.txt")
......@@ -40,14 +41,43 @@ def cv2_to_base64(image):
'utf8') #data.tostring()).decode('utf8')
for img_file in os.listdir(test_img_dir):
with open(os.path.join(test_img_dir, img_file), 'rb') as file:
def _check_image_file(path):
img_end = {'jpg', 'bmp', 'png', 'jpeg', 'rgb', 'tif', 'tiff', 'gif'}
return any([path.lower().endswith(e) for e in img_end])
test_img_list = []
if os.path.isfile(test_img_dir) and _check_image_file(test_img_dir):
test_img_list.append(test_img_dir)
elif os.path.isdir(test_img_dir):
for single_file in os.listdir(test_img_dir):
file_path = os.path.join(test_img_dir, single_file)
if os.path.isfile(file_path) and _check_image_file(file_path):
test_img_list.append(file_path)
if len(test_img_list) == 0:
raise Exception("not found any img file in {}".format(test_img_dir))
for img_file in test_img_list:
with open(img_file, 'rb') as file:
image_data = file.read()
image = cv2_to_base64(image_data)
res_list = []
fetch_map = client.predict(feed={"x": image}, fetch=[], batch=True)
one_batch_res = ocr_reader.postprocess(fetch_map, with_score=True)
if fetch_map is None:
print('no results')
else:
if "text" in fetch_map:
for x in fetch_map["text"]:
x = codecs.encode(x)
words = base64.b64decode(x).decode('utf-8')
res_list.append(words)
else:
try:
one_batch_res = ocr_reader.postprocess(
fetch_map, with_score=True)
for res in one_batch_res:
res_list.append(res[0])
except:
print('no results')
res = {"res": str(res_list)}
print(res)
deploy/pdserving/ocr_reader.py
浏览文件 @ 5016894e
......@@ -339,7 +339,7 @@ class CharacterOps(object):
class OCRReader(object):
def __init__(self,
algorithm="CRNN",
image_shape=[3, 32, 320],
image_shape=[3, 48, 320],
char_type="ch",
batch_num=1,
char_dict_path="./ppocr_keys_v1.txt"):
......@@ -356,7 +356,7 @@ class OCRReader(object):
def resize_norm_img(self, img, max_wh_ratio):
imgC, imgH, imgW = self.rec_image_shape
if self.character_type == "ch":
imgW = int(32 * max_wh_ratio)
imgW = int(imgH * max_wh_ratio)
h = img.shape[0]
w = img.shape[1]
ratio = w / float(h)
......@@ -377,7 +377,7 @@ class OCRReader(object):
def preprocess(self, img_list):
img_num = len(img_list)
norm_img_batch = []
max_wh_ratio = 0
max_wh_ratio = 320/48.
for ino in range(img_num):
h, w = img_list[ino].shape[0:2]
wh_ratio = w * 1.0 / h
......
deploy/pdserving/pipeline_http_client.py
浏览文件 @ 5016894e
......@@ -36,11 +36,27 @@ def cv2_to_base64(image):
return base64.b64encode(image).decode('utf8')
def _check_image_file(path):
img_end = {'jpg', 'bmp', 'png', 'jpeg', 'rgb', 'tif', 'tiff', 'gif'}
return any([path.lower().endswith(e) for e in img_end])
url = "http://127.0.0.1:9998/ocr/prediction"
test_img_dir = args.image_dir
for idx, img_file in enumerate(os.listdir(test_img_dir)):
with open(os.path.join(test_img_dir, img_file), 'rb') as file:
test_img_list = []
if os.path.isfile(test_img_dir) and _check_image_file(test_img_dir):
test_img_list.append(test_img_dir)
elif os.path.isdir(test_img_dir):
for single_file in os.listdir(test_img_dir):
file_path = os.path.join(test_img_dir, single_file)
if os.path.isfile(file_path) and _check_image_file(file_path):
test_img_list.append(file_path)
if len(test_img_list) == 0:
raise Exception("not found any img file in {}".format(test_img_dir))
for idx, img_file in enumerate(test_img_list):
with open(img_file, 'rb') as file:
image_data1 = file.read()
# print file name
print('{}{}{}'.format('*' * 10, img_file, '*' * 10))
......@@ -70,4 +86,4 @@ for idx, img_file in enumerate(os.listdir(test_img_dir)):
print(
"For details about error message, see PipelineServingLogs/pipeline.log"
)
print("==> total number of test imgs: ", len(os.listdir(test_img_dir)))
print("==> total number of test imgs: ", len(test_img_list))
deploy/pdserving/serving_client_conf.prototxt 0 → 100644
浏览文件 @ 5016894e
feed_var {
name: "x"
alias_name: "x"
is_lod_tensor: false
feed_type: 20
shape: 1
}
fetch_var {
name: "save_infer_model/scale_0.tmp_1"
alias_name: "save_infer_model/scale_0.tmp_1"
is_lod_tensor: false
fetch_type: 1
shape: 1
shape: 640
shape: 640
}
deploy/pdserving/web_service.py
浏览文件 @ 5016894e
......@@ -19,7 +19,7 @@ import copy
import cv2
import base64
# from paddle_serving_app.reader import OCRReader
from ocr_reader import OCRReader, DetResizeForTest
from ocr_reader import OCRReader, DetResizeForTest, ArgsParser
from paddle_serving_app.reader import Sequential, ResizeByFactor
from paddle_serving_app.reader import Div, Normalize, Transpose
from paddle_serving_app.reader import DBPostProcess, FilterBoxes, GetRotateCropImage, SortedBoxes
......@@ -63,7 +63,6 @@ class DetOp(Op):
dt_boxes_list = self.post_func(det_out, [ratio_list])
dt_boxes = self.filter_func(dt_boxes_list[0], [self.ori_h, self.ori_w])
out_dict = {"dt_boxes": dt_boxes, "image": self.raw_im}
return out_dict, None, ""
......@@ -86,7 +85,7 @@ class RecOp(Op):
dt_boxes = copy.deepcopy(self.dt_list)
feed_list = []
img_list = []
max_wh_ratio = 0
max_wh_ratio = 320 / 48.
## Many mini-batchs, the type of feed_data is list.
max_batch_size = 6 # len(dt_boxes)
......@@ -150,6 +149,7 @@ class RecOp(Op):
for i in range(dt_num):
text = rec_list[i]
dt_box = self.dt_list[i]
if text[1] >= 0.5:
result_list.append([text, dt_box.tolist()])
res = {"result": str(result_list)}
return res, None, ""
......@@ -163,5 +163,6 @@ class OcrService(WebService):
uci_service = OcrService(name="ocr")
uci_service.prepare_pipeline_config("config.yml")
FLAGS = ArgsParser().parse_args()
uci_service.prepare_pipeline_config(yml_dict=FLAGS.conf_dict)
uci_service.run_service()
doc/doc_ch/FAQ.md
浏览文件 @ 5016894e
......@@ -682,7 +682,7 @@ lr:
#### Q: 关于dygraph分支中，文本识别模型训练，要使用数据增强应该如何设置？
**A**：可以参考[配置文件](../../configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml)在Train['dataset']['transforms']添加RecAug字段，使数据增强生效。可以通过添加对aug_prob设置，表示每种数据增强采用的概率。aug_prob默认是0.4.由于tia数据增强特殊性，默认不采用，可以通过添加use_tia设置，使tia数据增强生效。详细设置可以参考[ISSUE 1744](https://github.com/PaddlePaddle/PaddleOCR/issues/1744)。
**A**：可以参考[配置文件](../../configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml)在Train['dataset']['transforms']添加RecAug字段，使数据增强生效。可以通过添加对aug_prob设置，表示每种数据增强采用的概率。aug_prob默认是0.4。详细设置可以参考[ISSUE 1744](https://github.com/PaddlePaddle/PaddleOCR/issues/1744)。
#### Q: 训练过程中，训练程序意外退出/挂起，应该如何解决？
......@@ -720,6 +720,13 @@ C++TensorRT预测需要使用支持TRT的预测库并在编译时打开[-DWITH_T
注：建议使用TensorRT大于等于6.1.0.5以上的版本。
#### Q: 为什么识别模型做预测的时候，预测图片的数量数量还会影响预测的精度
**A**： 推理时识别模型默认的batch_size=6, 如预测图片长度变化大，可能影响预测效果。如果出现上述问题可在推理的时候设置识别bs=1，命令如下：
```
python3 tools/infer/predict_rec.py --image_dir="./doc/imgs_words/ch/word_4.jpg" --rec_model_dir="./ch_PP-OCRv3_rec_infer/" --rec_batch_num=1
```
<a name="213"></a>
### 2.13 推理部署
......
doc/doc_ch/dataset/layout_datasets.md
浏览文件 @ 5016894e
......@@ -15,8 +15,8 @@
- **数据简介**：publaynet数据集的训练集合中包含35万张图像，验证集合中包含1.1万张图像。总共包含5个类别，分别是： `text, title, list, table, figure`。部分图像以及标注框可视化如下所示。
<div align="center">
<img src="../datasets/publaynet_demo/gt_PMC3724501_00006.jpg" width="500">
<img src="../datasets/publaynet_demo/gt_PMC5086060_00002.jpg" width="500">
<img src="../../datasets/publaynet_demo/gt_PMC3724501_00006.jpg" width="500">
<img src="../../datasets/publaynet_demo/gt_PMC5086060_00002.jpg" width="500">
</div>
- **下载地址**：https://developer.ibm.com/exchanges/data/all/publaynet/
......@@ -30,8 +30,8 @@
- **数据简介**：CDLA据集的训练集合中包含5000张图像，验证集合中包含1000张图像。总共包含10个类别，分别是： `Text, Title, Figure, Figure caption, Table, Table caption, Header, Footer, Reference, Equation`。部分图像以及标注框可视化如下所示。
<div align="center">
<img src="../datasets/CDLA_demo/val_0633.jpg" width="500">
<img src="../datasets/CDLA_demo/val_0941.jpg" width="500">
<img src="../../datasets/CDLA_demo/val_0633.jpg" width="500">
<img src="../../datasets/CDLA_demo/val_0941.jpg" width="500">
</div>
- **下载地址**：https://github.com/buptlihang/CDLA
......@@ -45,8 +45,8 @@
- **数据简介**：TableBank数据集包含Latex（训练集187199张，验证集7265张，测试集5719张）与Word（训练集73383张，验证集2735张，测试集2281张）两种类别的文档。仅包含`Table` 1个类别。部分图像以及标注框可视化如下所示。
<div align="center">
<img src="../datasets/tablebank_demo/004.png" height="700">
<img src="../datasets/tablebank_demo/005.png" height="700">
<img src="../../datasets/tablebank_demo/004.png" height="700">
<img src="../../datasets/tablebank_demo/005.png" height="700">
</div>
- **下载地址**：https://doc-analysis.github.io/tablebank-page/index.html
......
doc/doc_ch/detection.md
浏览文件 @ 5016894e
......@@ -13,6 +13,7 @@
- [2.5 分布式训练](#25-分布式训练)
- [2.6 知识蒸馏训练](#26-知识蒸馏训练)
- [2.7 其他训练环境](#27-其他训练环境)
- [2.8 模型微调](#28-模型微调)
- [3. 模型评估与预测](#3-模型评估与预测)
- [3.1 指标评估](#31-指标评估)
- [3.2 测试检测效果](#32-测试检测效果)
......@@ -141,7 +142,8 @@ python3 tools/train.py -c configs/det/det_mv3_db.yml \
Global.use_amp=True Global.scale_loss=1024.0 Global.use_dynamic_loss_scaling=True
```
<a name="26---fleet---"></a>
<a name="25---fleet---"></a>
## 2.5 分布式训练
多机多卡训练时，通过 `--ips` 参数设置使用的机器IP地址，通过 `--gpus` 参数设置使用的GPU ID：
......@@ -151,7 +153,7 @@ 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
```
**注意:** 采用多机多卡训练时，需要替换上面命令中的ips值为您机器的地址，机器之间需要能够相互ping通。另外，训练时需要在多个机器上分别启动命令。查看机器ip地址的命令为`ifconfig`。
**注意:** （1）采用多机多卡训练时，需要替换上面命令中的ips值为您机器的地址，机器之间需要能够相互ping通；（2）训练时需要在多个机器上分别启动命令。查看机器ip地址的命令为`ifconfig`；（3）更多关于分布式训练的性能优势等信息，请参考：[分布式训练教程](./distributed_training.md)。
<a name="26---distill---"></a>
......@@ -177,6 +179,13 @@ Windows平台只支持`单卡`的训练与预测，指定GPU进行训练`set CUD
- Linux DCU
DCU设备上运行需要设置环境变量 `export HIP_VISIBLE_DEVICES=0,1,2,3`，其余训练评估预测命令与Linux GPU完全相同。
<a name="28-模型微调"></a>
## 2.8 模型微调
实际使用过程中，建议加载官方提供的预训练模型，在自己的数据集中进行微调，关于检测模型的微调方法，请参考：[模型微调教程](./finetune.md)。
<a name="3--------"></a>
# 3. 模型评估与预测
......@@ -196,6 +205,7 @@ python3 tools/eval.py -c configs/det/det_mv3_db.yml -o Global.checkpoints="{pat
## 3.2 测试检测效果
测试单张图像的检测效果：
```shell
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy"
```
......@@ -226,14 +236,19 @@ python3 tools/export_model.py -c configs/det/det_mv3_db.yml -o Global.pretrained
```
DB检测模型inference 模型预测：
```shell
python3 tools/infer/predict_det.py --det_algorithm="DB" --det_model_dir="./output/det_db_inference/" --image_dir="./doc/imgs/" --use_gpu=True
```
如果是其他检测，比如EAST模型，det_algorithm参数需要修改为EAST，默认为DB算法：
```shell
python3 tools/infer/predict_det.py --det_algorithm="EAST" --det_model_dir="./output/det_db_inference/" --image_dir="./doc/imgs/" --use_gpu=True
```
更多关于推理超参数的配置与解释，请参考：[模型推理超参数解释教程](./inference_args.md)。
<a name="5-faq"></a>
# 5. FAQ
......
doc/doc_ch/distributed_training.md
浏览文件 @ 5016894e
......@@ -41,11 +41,16 @@ python3 -m paddle.distributed.launch \
## 性能效果测试
* 基于单机8卡P40，和2机8卡P40，在26W公开识别数据集(LSVT, RCTW, MTWI)上进行训练，最终耗时如下。
* 在2机8卡P40的机器上，基于26W公开识别数据集(LSVT, RCTW, MTWI)上进行训练，最终耗时如下。
| 模型 | 配置文件 | 机器数量 | 每台机器的GPU数量 | 训练时间 | 识别Acc | 加速比 |
| :----------------------: | :------------: | :------------: | :---------------: | :----------: | :-----------: | :-----------: |
| 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% |
| 模型 | 配置 | 精度 | 单机8卡耗时 | 2机8卡耗时 | 加速比 |
|------|-----|--------|--------|--------|-----|
| 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** |
可以看出，精度没有下降的情况下，训练时间由60h缩短为了40h，加速比可以达到60h/40h=150%，效率为60h/(40h*2)=75%。
* 在4机8卡V100的机器上，基于全量数据训练，最终耗时如下
| 模型 | 配置 | 精度 | 单机8卡耗时 | 4机8卡耗时 | 加速比 |
|------|-----|--------|--------|--------|-----|
| SVTR | [ch_PP-OCRv3_rec_distillation.yml](../../configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml) | 74.0% | 10d | 2.84d | **3.5** |
doc/doc_ch/inference_args.md 0 → 100644
浏览文件 @ 5016894e
# PaddleOCR模型推理参数解释
在使用PaddleOCR进行模型推理时，可以自定义修改参数，来修改模型、数据、预处理、后处理等内容（参数文件：[utility.py](../../tools/infer/utility.py)），详细的参数解释如下所示。
* 全局信息
| 参数名称 | 类型 | 默认值 | 含义 |
| :--: | :--: | :--: | :--: |
| image_dir | str | 无，必须显式指定 | 图像或者文件夹路径 |
| vis_font_path | str | "./doc/fonts/simfang.ttf" | 用于可视化的字体路径 |
| drop_score | float | 0.5 | 识别得分小于该值的结果会被丢弃，不会作为返回结果 |
| use_pdserving | bool | False | 是否使用Paddle Serving进行预测 |
| warmup | bool | False | 是否开启warmup，在统计预测耗时的时候，可以使用这种方法 |
| draw_img_save_dir | str | "./inference_results" | 系统串联预测OCR结果的保存文件夹 |
| save_crop_res | bool | False | 是否保存OCR的识别文本图像 |
| crop_res_save_dir | str | "./output" | 保存OCR识别出来的文本图像路径 |
| use_mp | bool | False | 是否开启多进程预测 |
| total_process_num | int | 6 | 开启的进城数，`use_mp`为`True`时生效 |
| process_id | int | 0 | 当前进程的id号，无需自己修改 |
| benchmark | bool | False | 是否开启benchmark，对预测速度、显存占用等进行统计 |
| save_log_path | str | "./log_output/" | 开启`benchmark`时，日志结果的保存文件夹 |
| show_log | bool | True | 是否显示预测中的日志信息 |
| use_onnx | bool | False | 是否开启onnx预测 |
* 预测引擎相关
| 参数名称 | 类型 | 默认值 | 含义 |
| :--: | :--: | :--: | :--: |
| use_gpu | bool | True | 是否使用GPU进行预测 |
| ir_optim | bool | True | 是否对计算图进行分析与优化，开启后可以加速预测过程 |
| use_tensorrt | bool | False | 是否开启tensorrt |
| min_subgraph_size | int | 15 | tensorrt中最小子图size，当子图的size大于该值时，才会尝试对该子图使用trt engine计算 |
| precision | str | fp32 | 预测的精度，支持`fp32`, `fp16`, `int8` 3种输入 |
| enable_mkldnn | bool | True | 是否开启mkldnn |
| cpu_threads | int | 10 | 开启mkldnn时，cpu预测的线程数 |
* 文本检测模型相关
| 参数名称 | 类型 | 默认值 | 含义 |
| :--: | :--: | :--: | :--: |
| det_algorithm | str | "DB" | 文本检测算法名称，目前支持`DB`, `EAST`, `SAST`, `PSE` |
| det_model_dir | str | xx | 检测inference模型路径 |
| det_limit_side_len | int | 960 | 检测的图像边长限制 |
| det_limit_type | str | "max" | 检测的变成限制类型，目前支持`min`, `max`，`min`表示保证图像最短边不小于`det_limit_side_len`，`max`表示保证图像最长边不大于`det_limit_side_len` |
其中，DB算法相关参数如下
| 参数名称 | 类型 | 默认值 | 含义 |
| :--: | :--: | :--: | :--: |
| det_db_thresh | float | 0.3 | DB输出的概率图中，得分大于该阈值的像素点才会被认为是文字像素点 |
| det_db_box_thresh | float | 0.6 | 检测结果边框内，所有像素点的平均得分大于该阈值时，该结果会被认为是文字区域 |
| det_db_unclip_ratio | float | 1.5 | `Vatti clipping`算法的扩张系数，使用该方法对文字区域进行扩张 |
| max_batch_size | int | 10 | 预测的batch size |
| use_dilation | bool | False | 是否对分割结果进行膨胀以获取更优检测效果 |
| det_db_score_mode | str | "fast" | DB的检测结果得分计算方法，支持`fast`和`slow`，`fast`是根据polygon的外接矩形边框内的所有像素计算平均得分，`slow`是根据原始polygon内的所有像素计算平均得分，计算速度相对较慢一些，但是更加准确一些。 |
EAST算法相关参数如下
| 参数名称 | 类型 | 默认值 | 含义 |
| :--: | :--: | :--: | :--: |
| det_east_score_thresh | float | 0.8 | EAST后处理中score map的阈值 |
| det_east_cover_thresh | float | 0.1 | EAST后处理中文本框的平均得分阈值 |
| det_east_nms_thresh | float | 0.2 | EAST后处理中nms的阈值 |
SAST算法相关参数如下
| 参数名称 | 类型 | 默认值 | 含义 |
| :--: | :--: | :--: | :--: |
| det_sast_score_thresh | float | 0.5 | SAST后处理中的得分阈值 |
| det_sast_nms_thresh | float | 0.5 | SAST后处理中nms的阈值 |
| det_sast_polygon | bool | False | 是否多边形检测，弯曲文本场景（如Total-Text）设置为True |
PSE算法相关参数如下
| 参数名称 | 类型 | 默认值 | 含义 |
| :--: | :--: | :--: | :--: |
| det_pse_thresh | float | 0.0 | 对输出图做二值化的阈值 |
| det_pse_box_thresh | float | 0.85 | 对box进行过滤的阈值，低于此阈值的丢弃 |
| det_pse_min_area | float | 16 | box的最小面积，低于此阈值的丢弃 |
| det_pse_box_type | str | "box" | 返回框的类型，box:四点坐标，poly: 弯曲文本的所有点坐标 |
| det_pse_scale | int | 1 | 输入图像相对于进后处理的图的比例，如`640*640`的图像，网络输出为`160*160`，scale为2的情况下，进后处理的图片shape为`320*320`。这个值调大可以加快后处理速度，但是会带来精度的下降 |
* 文本识别模型相关
| 参数名称 | 类型 | 默认值 | 含义 |
| :--: | :--: | :--: | :--: |
| rec_algorithm | str | "CRNN" | 文本识别算法名称，目前支持`CRNN`, `SRN`, `RARE`, `NETR`, `SAR` |
| rec_model_dir | str | 无，如果使用识别模型，该项是必填项 | 识别inference模型路径 |
| rec_image_shape | list | [3, 32, 320] | 识别时的图像尺寸， |
| rec_batch_num | int | 6 | 识别的batch size |
| max_text_length | int | 25 | 识别结果最大长度，在`SRN`中有效 |
| rec_char_dict_path | str | "./ppocr/utils/ppocr_keys_v1.txt" | 识别的字符字典文件 |
| use_space_char | bool | True | 是否包含空格，如果为`True`，则会在最后字符字典中补充`空格`字符 |
* 端到端文本检测与识别模型相关
| 参数名称 | 类型 | 默认值 | 含义 |
| :--: | :--: | :--: | :--: |
| e2e_algorithm | str | "PGNet" | 端到端算法名称，目前支持`PGNet` |
| e2e_model_dir | str | 无，如果使用端到端模型，该项是必填项 | 端到端模型inference模型路径 |
| e2e_limit_side_len | int | 768 | 端到端的输入图像边长限制 |
| e2e_limit_type | str | "max" | 端到端的边长限制类型，目前支持`min`, `max`，`min`表示保证图像最短边不小于`e2e_limit_side_len`，`max`表示保证图像最长边不大于`e2e_limit_side_len` |
| e2e_pgnet_score_thresh | float | 0.5 | 端到端得分阈值，小于该阈值的结果会被丢弃 |
| e2e_char_dict_path | str | "./ppocr/utils/ic15_dict.txt" | 识别的字典文件路径 |
| e2e_pgnet_valid_set | str | "totaltext" | 验证集名称，目前支持`totaltext`, `partvgg`，不同数据集对应的后处理方式不同，与训练过程保持一致即可 |
| e2e_pgnet_mode | str | "fast" | PGNet的检测结果得分计算方法，支持`fast`和`slow`，`fast`是根据polygon的外接矩形边框内的所有像素计算平均得分，`slow`是根据原始polygon内的所有像素计算平均得分，计算速度相对较慢一些，但是更加准确一些。 |
* 方向分类器模型相关
| 参数名称 | 类型 | 默认值 | 含义 |
| :--: | :--: | :--: | :--: |
| use_angle_cls | bool | False | 是否使用方向分类器 |
| cls_model_dir | str | 无，如果需要使用，则必须显式指定路径 | 方向分类器inference模型路径 |
| cls_image_shape | list | [3, 48, 192] | 预测尺度 |
| label_list | list | ['0', '180'] | class id对应的角度值 |
| cls_batch_num | int | 6 | 方向分类器预测的batch size |
| cls_thresh | float | 0.9 | 预测阈值，模型预测结果为180度，且得分大于该阈值时，认为最终预测结果为180度，需要翻转 |
doc/doc_ch/ppocr_introduction.md
浏览文件 @ 5016894e
......@@ -30,11 +30,11 @@ PP-OCR系统pipeline如下：
PP-OCR系统在持续迭代优化，目前已发布PP-OCR和PP-OCRv2两个版本：
PP-OCR从骨干网络选择和调整、预测头部的设计、数据增强、学习率变换策略、正则化参数选择、预训练模型使用以及模型自动裁剪量化8个方面，采用19个有效策略，对各个模块的模型进行效果调优和瘦身(如绿框所示)，最终得到整体大小为3.5M的超轻量中英文OCR和2.8M的英文数字OCR。更多细节请参考PP-OCR技术方案 https://arxiv.org/abs/2009.09941
PP-OCR从骨干网络选择和调整、预测头部的设计、数据增强、学习率变换策略、正则化参数选择、预训练模型使用以及模型自动裁剪量化8个方面，采用19个有效策略，对各个模块的模型进行效果调优和瘦身(如绿框所示)，最终得到整体大小为3.5M的超轻量中英文OCR和2.8M的英文数字OCR。更多细节请参考[PP-OCR技术报告](https://arxiv.org/abs/2009.09941)。
#### PP-OCRv2
PP-OCRv2在PP-OCR的基础上，进一步在5个方面重点优化，检测模型采用CML协同互学习知识蒸馏策略和CopyPaste数据增广策略；识别模型采用LCNet轻量级骨干网络、UDML 改进知识蒸馏策略和[Enhanced CTC loss](./enhanced_ctc_loss.md)损失函数改进（如上图红框所示），进一步在推理速度和预测效果上取得明显提升。更多细节请参考PP-OCRv2[技术报告](https://arxiv.org/abs/2109.03144)。
PP-OCRv2在PP-OCR的基础上，进一步在5个方面重点优化，检测模型采用CML协同互学习知识蒸馏策略和CopyPaste数据增广策略；识别模型采用LCNet轻量级骨干网络、UDML 改进知识蒸馏策略和[Enhanced CTC loss](./enhanced_ctc_loss.md)损失函数改进（如上图红框所示），进一步在推理速度和预测效果上取得明显提升。更多细节请参考[PP-OCRv2技术报告](https://arxiv.org/abs/2109.03144)。
#### PP-OCRv3
......@@ -48,7 +48,7 @@ PP-OCRv3系统pipeline如下：
<img src="../ppocrv3_framework.png" width="800">
</div>
更多细节请参考PP-OCRv3[技术报告](./PP-OCRv3_introduction.md)。
更多细节请参考[PP-OCRv3技术报告](https://arxiv.org/abs/2206.03001v2) 👉[中文简洁版](./PP-OCRv3_introduction.md)
<a name="2"></a>
......
doc/doc_ch/quickstart.md
浏览文件 @ 5016894e
......@@ -101,8 +101,17 @@ cd /path/to/ppocr_img
['韩国小馆', 0.994467]
```
**版本说明**
paddleocr默认使用PP-OCRv3模型(`--ocr_version PP-OCRv3`)，如需使用其他版本可通过设置参数`--ocr_version`，具体版本说明如下：
| 版本名称 | 版本说明 |
| --- | --- |
| PP-OCRv3 | 支持中、英文检测和识别，方向分类器，支持多语种识别 |
| PP-OCRv2 | 支持中英文的检测和识别，方向分类器，多语言暂未更新 |
| PP-OCR | 支持中、英文检测和识别，方向分类器，支持多语种识别 |
如需使用2.0模型，请指定参数`--ocr_version PP-OCR`，paddleocr默认使用PP-OCRv3模型(`--ocr_version PP-OCRv3`)。更多whl包使用可参考[whl包文档](./whl.md)
如需新增自己训练的模型，可以在[paddleocr](../../paddleocr.py)中增加模型链接和字段，重新编译即可。
更多whl包使用可参考[whl包文档](./whl.md)
<a name="212"></a>
......
doc/doc_ch/recognition.md
浏览文件 @ 5016894e
......@@ -18,6 +18,7 @@
- [2.6. 知识蒸馏训练](#26-知识蒸馏训练)
- [2.7. 多语言模型训练](#27-多语言模型训练)
- [2.8. 其他训练环境](#28-其他训练环境)
- [2.9. 模型微调](#29-模型微调)
- [3. 模型评估与预测](#3-模型评估与预测)
- [3.1. 指标评估](#31-指标评估)
- [3.2. 测试识别效果](#32-测试识别效果)
......@@ -217,6 +218,30 @@ python3 tools/train.py -c configs/rec/PP-OCRv3/en_PP-OCRv3_rec.yml -o Global.pre
python3 -m paddle.distributed.launch --gpus '0,1,2,3' tools/train.py -c configs/rec/PP-OCRv3/en_PP-OCRv3_rec.yml -o Global.pretrained_model=./pretrain_models/en_PP-OCRv3_rec_train/best_accuracy
```
正常启动训练后，会看到以下log输出：
```
[2022/02/22 07:58:05] root INFO: epoch: [1/800], iter: 10, lr: 0.000000, loss: 0.754281, acc: 0.000000, norm_edit_dis: 0.000008, reader_cost: 0.55541 s, batch_cost: 0.91654 s, samples: 1408, ips: 153.62133
[2022/02/22 07:58:13] root INFO: epoch: [1/800], iter: 20, lr: 0.000001, loss: 0.924677, acc: 0.000000, norm_edit_dis: 0.000008, reader_cost: 0.00236 s, batch_cost: 0.28528 s, samples: 1280, ips: 448.68599
[2022/02/22 07:58:23] root INFO: epoch: [1/800], iter: 30, lr: 0.000002, loss: 0.967231, acc: 0.000000, norm_edit_dis: 0.000008, reader_cost: 0.14527 s, batch_cost: 0.42714 s, samples: 1280, ips: 299.66507
[2022/02/22 07:58:31] root INFO: epoch: [1/800], iter: 40, lr: 0.000003, loss: 0.895318, acc: 0.000000, norm_edit_dis: 0.000008, reader_cost: 0.00173 s, batch_cost: 0.27719 s, samples: 1280, ips: 461.77252
```
log 中自动打印如下信息：
| 字段 | 含义 |
| :----: | :------: |
| epoch | 当前迭代轮次 |
| iter | 当前迭代次数 |
| lr | 当前学习率 |
| loss | 当前损失函数 |
| acc | 当前batch的准确率 |
| norm_edit_dis | 当前 batch 的编辑距离 |
| reader_cost | 当前 batch 数据处理耗时 |
| batch_cost | 当前 batch 总耗时 |
| samples | 当前 batch 内的样本数 |
| ips | 每秒处理图片的数量 |
PaddleOCR支持训练和评估交替进行, 可以在 `configs/rec/PP-OCRv3/en_PP-OCRv3_rec.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/en_PP-OCRv3_rec/best_accuracy` 。
......@@ -363,7 +388,7 @@ python3 -m paddle.distributed.launch --ips="xx.xx.xx.xx,xx.xx.xx.xx" --gpus '0,1
-o Global.pretrained_model=./pretrain_models/en_PP-OCRv3_rec_train/best_accuracy
```
**注意:** 采用多机多卡训练时，需要替换上面命令中的ips值为您机器的地址，机器之间需要能够相互ping通。另外，训练时需要在多个机器上分别启动命令。查看机器ip地址的命令为`ifconfig`。
**注意:** （1）采用多机多卡训练时，需要替换上面命令中的ips值为您机器的地址，机器之间需要能够相互ping通；（2）训练时需要在多个机器上分别启动命令。查看机器ip地址的命令为`ifconfig`；（3）更多关于分布式训练的性能优势等信息，请参考：[分布式训练教程](./distributed_training.md)。
## 2.6. 知识蒸馏训练
......@@ -438,6 +463,11 @@ Windows平台只支持`单卡`的训练与预测，指定GPU进行训练`set CUD
- Linux DCU
DCU设备上运行需要设置环境变量 `export HIP_VISIBLE_DEVICES=0,1,2,3`，其余训练评估预测命令与Linux GPU完全相同。
## 2.9 模型微调
实际使用过程中，建议加载官方提供的预训练模型，在自己的数据集中进行微调，关于识别模型的微调方法，请参考：[模型微调教程](./finetune.md)。
# 3. 模型评估与预测
## 3.1. 指标评估
......@@ -540,12 +570,13 @@ inference/en_PP-OCRv3_rec/
- 自定义模型推理
如果训练时修改了文本的字典，在使用inference模型预测时，需要通过`--rec_char_dict_path`指定使用的字典路径
如果训练时修改了文本的字典，在使用inference模型预测时，需要通过`--rec_char_dict_path`指定使用的字典路径，更多关于推理超参数的配置与解释，请参考：[模型推理超参数解释教程](./inference_args.md)。
```
python3 tools/infer/predict_rec.py --image_dir="./doc/imgs_words_en/word_336.png" --rec_model_dir="./your inference model" --rec_image_shape="3, 48, 320" --rec_char_dict_path="your text dict path"
```
# 5. FAQ
Q1: 训练模型转inference 模型之后预测效果不一致？
......
doc/doc_en/PP-OCRv3_introduction_en.md
浏览文件 @ 5016894e
......@@ -76,7 +76,7 @@ LK-PAN (Large Kernel PAN) is a lightweight [PAN](https://arxiv.org/pdf/1803.0153
**(2) DML: Deep Mutual Learning Strategy for Teacher Model**
[DML](https://arxiv.org/abs/1706.00384)(Collaborative Mutual Learning), as shown in the figure below, can effectively improve the accuracy of the text detection model by learning from each other with two models with the same structure. The DML strategy is adopted in the teacher model training, and the hmean is increased from 85% to 86%. By updating the teacher model of CML in PP-OCRv2 to the above-mentioned higher-precision one, the hmean of the student model can be further improved from 83.2% to 84.3%.
[DML](https://arxiv.org/abs/1706.00384)(Deep Mutual Learning), as shown in the figure below, can effectively improve the accuracy of the text detection model by learning from each other with two models with the same structure. The DML strategy is adopted in the teacher model training, and the hmean is increased from 85% to 86%. By updating the teacher model of CML in PP-OCRv2 to the above-mentioned higher-precision one, the hmean of the student model can be further improved from 83.2% to 84.3%.
<div align="center">
......@@ -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.
<div align="center">
<img src="../ppocr_v3/v3_rec_pipeline.png" width=800>
......
doc/doc_en/algorithm_rec_starnet.md 0 → 100644
浏览文件 @ 5016894e
# 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)
<a name="1"></a>
## 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)|
<a name="2"></a>
## 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.
<a name="3"></a>
## 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:
<a name="3-1"></a>
### 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
````
<a name="3-2"></a>
### 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
````
<a name="3-3"></a>
### 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
````
<a name="4"></a>
## 4. Inference
<a name="4-1"></a>
### 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"
````
![](../imgs_words_en/word_336.png)
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)
```
<a name="4-2"></a>
### 4.2 C++ Inference
After preparing the inference model, refer to the [cpp infer](../../deploy/cpp_infer/) tutorial to operate.
<a name="4-3"></a>
### 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.
<a name="4-4"></a>
### 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.
<a name="5"></a>
## 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}
}
```
doc/doc_en/detection_en.md
浏览文件 @ 5016894e
......@@ -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
......
doc/doc_en/distributed_training.md → doc/doc_en/distributed_training_en.md
浏览文件 @ 5016894e
......@@ -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** |
doc/doc_en/ppocr_introduction_en.md
浏览文件 @ 5016894e
......@@ -29,10 +29,10 @@ 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
......@@ -46,7 +46,7 @@ PP-OCRv3 pipeline is as follows:
<img src="../ppocrv3_framework.png" width="800">
</div>
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).
<a name="2"></a>
## 2. Features
......
doc/doc_en/quickstart_en.md
浏览文件 @ 5016894e
......@@ -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)
<a name="212-multi-language-model"></a>
#### 2.1.2 Multi-language Model
......
doc/doc_en/recognition_en.md
浏览文件 @ 5016894e
......@@ -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).
<a name="kd"></a>
### 2.6 Training with Knowledge Distillation
......
paddleocr.py
浏览文件 @ 5016894e
......@@ -154,8 +154,14 @@ 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': {
'ch': {
......
ppocr/data/imaug/__init__.py
浏览文件 @ 5016894e
......@@ -22,7 +22,7 @@ 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, \
from .rec_img_aug import BaseDataAugmentation, RecAug, RecConAug, RecResizeImg, ClsResizeImg, \
SRNRecResizeImg, NRTRRecResizeImg, SARRecResizeImg, PRENResizeImg
from .ssl_img_aug import SSLRotateResize
from .randaugment import RandAugment
......
ppocr/data/imaug/rec_img_aug.py
浏览文件 @ 5016894e
......@@ -22,13 +22,74 @@ from .text_image_aug import tia_perspective, tia_stretch, tia_distort
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
......@@ -359,7 +420,7 @@ def flag():
return 1 if random.random() > 0.5000001 else -1
def cvtColor(img):
def hsv_aug(img):
"""
cvtColor
"""
......@@ -427,50 +488,6 @@ 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 +571,3 @@ 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
ppocr/data/simple_dataset.py
浏览文件 @ 5016894e
......@@ -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)
......
ppocr/losses/rec_aster_loss.py
浏览文件 @ 5016894e
......@@ -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(
......
ppocr/losses/table_att_loss.py
浏览文件 @ 5016894e
......@@ -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
......
ppocr/modeling/backbones/kie_unet_sdmgr.py
浏览文件 @ 5016894e
......@@ -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]
ppocr/modeling/backbones/rec_resnet_fpn.py
浏览文件 @ 5016894e
......@@ -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
......
ppocr/modeling/heads/rec_srn_head.py
浏览文件 @ 5016894e
......@@ -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
......
ppocr/modeling/heads/self_attention.py
浏览文件 @ 5016894e
......@@ -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)
......
ppstructure/table/README.md
浏览文件 @ 5016894e
......@@ -18,7 +18,7 @@ The table recognition mainly contains three models
The table recognition flow chart is as follows
![tableocr_pipeline](../../doc/table/tableocr_pipeline_en.jpg)
![tableocr_pipeline](../docs/table/tableocr_pipeline_en.jpg)
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.
......
ppstructure/table/predict_table.py
浏览文件 @ 5016894e
......@@ -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__":
......
test_tipc/build_server.sh 0 → 100644
浏览文件 @ 5016894e
#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 ../../
test_tipc/common_func.sh
浏览文件 @ 5016894e
......@@ -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
}
test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_det/train_infer_python.txt
浏览文件 @ 5016894e
===========================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
......
test_tipc/configs/ch_PP-OCRv2_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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]}]
test_tipc/configs/ch_PP-OCRv2_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================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
......
test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================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/
......
test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_det_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_det_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_det_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_det_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_det_PACT/train_infer_python.txt
浏览文件 @ 5016894e
===========================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
......
test_tipc/configs/ch_PP-OCRv2_det_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================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: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
......
test_tipc/configs/ch_PP-OCRv2_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_rec/train_infer_python.txt
浏览文件 @ 5016894e
===========================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
......
test_tipc/configs/ch_PP-OCRv2_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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]}]
test_tipc/configs/ch_PP-OCRv2_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================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
......
test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================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
......
test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_rec_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_rec_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_rec_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_rec_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv2_rec_PACT/train_infer_python.txt
浏览文件 @ 5016894e
===========================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
......
test_tipc/configs/ch_PP-OCRv2_rec_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================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:amp
Global.epoch_num:lite_train_lite_infer=3|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:null
Global.pretrained_model:pretrain_models/ch_PP-OCRv2_rec_train/best_accuracy
train_model_name:latest
train_infer_img_dir:./inference/rec_inference
null:null
......@@ -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
......
test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================cpp_infer_params===========================
model_name:ocr_system_v3
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: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/
--rec_model_dir:./inference/ch_PP-OCRv3_rec_infer/
......
test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
......@@ -6,10 +6,10 @@ 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|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/
......
test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================cpp_infer_params===========================
model_name:ocr_det_v3
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: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
......
test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
浏览文件 @ 5016894e
===========================paddle2onnx_params===========================
model_name:ocr_det_v3
model_name:ch_PP-OCRv3_det
python:python3.7
2onnx: paddle2onnx
--model_dir:./inference/ch_PP-OCRv3_det_infer/
--det_model_dir:./inference/ch_PP-OCRv3_det_infer/
--model_filename:inference.pdmodel
--params_filename:inference.pdiparams
--save_file:./inference/det_v3_onnx/model.onnx
--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
test_tipc/configs/ch_PP-OCRv3_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================serving_params===========================
model_name:ocr_det_v3
model_name:ch_PP-OCRv3_det
python:python3.7
trans_model:-m paddle_serving_client.convert
--dirname:./inference/ch_PP-OCRv3_det_infer/
--det_dirname:./inference/ch_PP-OCRv3_det_infer/
--model_filename:inference.pdmodel
--params_filename:inference.pdiparams
--serving_server:./deploy/pdserving/ppocr_det_v3_serving/
--serving_client:./deploy/pdserving/ppocr_det_v3_client/
--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: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.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
test_tipc/configs/ch_PP-OCRv3_det/train_infer_python.txt
浏览文件 @ 5016894e
===========================train_params===========================
model_name:ch_PPOCRv3_det
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=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
......
test_tipc/configs/ch_PP-OCRv3_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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]}]
test_tipc/configs/ch_PP-OCRv3_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================train_params===========================
model_name:ch_PPOCRv3_det
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=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
......
test_tipc/configs/ch_PP-OCRv3_det_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3_det_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================kl_quant_params===========================
model_name:PPOCRv3_ocr_det_kl
model_name:ch_PP-OCRv3_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-OCRv3/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/
......
test_tipc/configs/ch_PP-OCRv3_det_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3_det_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3_det_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3_det_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3_det_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3_det_PACT/train_infer_python.txt
浏览文件 @ 5016894e
===========================train_params===========================
model_name:ch_PPOCRv3_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: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
......
test_tipc/configs/ch_PP-OCRv3_det_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================train_params===========================
model_name:ch_PPOCRv3_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.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
......
test_tipc/configs/ch_PP-OCRv3_rec/ch_PP-OCRv3_rec_distillation.yml
浏览文件 @ 5016894e
......@@ -153,7 +153,7 @@ Train:
data_dir: ./train_data/ic15_data/
ext_op_transform_idx: 1
label_file_list:
- ./train_data/ic15_data/rec_gt_train.txt
- ./train_data/ic15_data/rec_gt_train_lite.txt
transforms:
- DecodeImage:
img_mode: BGR
......@@ -183,7 +183,7 @@ Eval:
name: SimpleDataSet
data_dir: ./train_data/ic15_data
label_file_list:
- ./train_data/ic15_data/rec_gt_test.txt
- ./train_data/ic15_data/rec_gt_test_lite.txt
transforms:
- DecodeImage:
img_mode: BGR
......
test_tipc/configs/ch_PP-OCRv3_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================cpp_infer_params===========================
model_name:ocr_rec_v3
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:True|False
--cpu_threads:1|6
--enable_mkldnn:False
--cpu_threads:6
--rec_batch_num:6
--use_tensorrt:False|True
--precision:fp32|fp16
--use_tensorrt:False
--precision:fp32
--rec_model_dir:
--image_dir:./inference/rec_inference/
null:null
......
test_tipc/configs/ch_PP-OCRv3_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
浏览文件 @ 5016894e
===========================paddle2onnx_params===========================
model_name:ocr_rec_v3
model_name:ch_PP-OCRv3_rec
python:python3.7
2onnx: paddle2onnx
--model_dir:./inference/ch_PP-OCRv3_rec_infer/
--det_model_dir:
--model_filename:inference.pdmodel
--params_filename:inference.pdiparams
--save_file:./inference/rec_v3_onnx/model.onnx
--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
--image_dir:./inference/rec_inference/
\ No newline at end of file
test_tipc/configs/ch_PP-OCRv3_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================serving_params===========================
model_name:ocr_rec_v3
model_name:ch_PP-OCRv3_rec
python:python3.7
trans_model:-m paddle_serving_client.convert
--dirname:./inference/ch_PP-OCRv3_rec_infer/
--det_dirname:null
--model_filename:inference.pdmodel
--params_filename:inference.pdiparams
--serving_server:./deploy/pdserving/ppocr_rec_v3_serving/
--serving_client:./deploy/pdserving/ppocr_rec_v3_client/
--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.rec.local_service_conf.devices:gpu|null
op.rec.local_service_conf.use_mkldnn: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
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
test_tipc/configs/ch_PP-OCRv3_rec/train_infer_python.txt
浏览文件 @ 5016894e
===========================train_params===========================
model_name:PPOCRv3_ocr_rec
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=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 --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
......
test_tipc/configs/ch_PP-OCRv3_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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]}]
test_tipc/configs/ch_PP-OCRv3_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================train_params===========================
model_name:PPOCRv3_ocr_rec
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=3|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:null
......@@ -39,11 +39,11 @@ infer_export:null
infer_quant:False
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
null:null
......
test_tipc/configs/ch_PP-OCRv3_rec_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3_rec_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================kl_quant_params===========================
model_name:PPOCRv3_ocr_rec_kl
model_name:ch_PP-OCRv3_rec_KL
python:python3.7
Global.pretrained_model:null
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|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
......
test_tipc/configs/ch_PP-OCRv3_rec_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3_rec_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3_rec_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3_rec_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3_rec_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_PP-OCRv3_rec_PACT/train_infer_python.txt
浏览文件 @ 5016894e
===========================train_params===========================
model_name:ch_PPOCRv3_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: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-OCRv3_rec_train/best_accuracy
......@@ -34,16 +34,16 @@ distill_export:null
export1:null
export2:null
inference_dir:Student
infer_model:./inference/ch_PP-OCRv3_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 --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
......
test_tipc/configs/ch_PP-OCRv3_rec_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================train_params===========================
model_name:ch_PPOCRv3_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.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
......@@ -39,11 +39,11 @@ infer_export:null
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
null:null
......
test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================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
......
test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_infer_python_jetson.txt
浏览文件 @ 5016894e
===========================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
......
test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================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:gpu|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.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
test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_linux_gpu_fleet_amp_infer_python_linux_gpu_cpu.txt → test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================train_params===========================
model_name:ocr_det
model_name:ch_ppocr_mobile_v2.0_det
python:python3.7
gpu_list:xx.xx.xx.xx,yy.yy.yy.yy;0,1
gpu_list:192.168.0.1,192.168.0.2;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.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
......@@ -12,10 +12,10 @@ 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
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
......@@ -26,10 +26,10 @@ 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
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
......@@ -38,14 +38,16 @@ 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
--use_gpu:False
--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
--benchmark:True
null:null
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,640,640]}];[{float32,[3,960,960]}]
\ No newline at end of file
test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_mac_cpu_normal_normal_infer_python_mac_cpu.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/ch_ppocr_mobile_v2.0_det/train_windows_gpu_normal_normal_infer_python_windows_cpu_gpu.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/ch_ppocr_mobile_v2.0_det_FPGM/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/ch_ppocr_mobile_v2.0_det_FPGM/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_det_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/ch_ppocr_mobile_v2.0_det_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
......@@ -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=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
......
test_tipc/configs/ch_ppocr_mobile_v2.0_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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]}]
test_tipc/configs/ch_ppocr_mobile_v2.0_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/ch_ppocr_mobile_v2.0_rec_FPGM/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/ch_ppocr_mobile_v2.0_rec_FPGM/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_rec_KL/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/ch_ppocr_mobile_v2.0_rec_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
......@@ -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.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/
......
test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_serving_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_server_v2.0/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_server_v2.0_det/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_server_v2.0_det/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_server_v2.0_det/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_server_v2.0_det/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/ch_ppocr_server_v2.0_det/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_server_v2.0_det/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/ch_ppocr_server_v2.0_rec/model_linux_gpu_normal_normal_infer_cpp_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_server_v2.0_rec/model_linux_gpu_normal_normal_paddle2onnx_python_linux_cpu.txt
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_server_v2.0_rec/model_linux_gpu_normal_normal_serving_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
===========================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
test_tipc/configs/ch_ppocr_server_v2.0_rec/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/ch_ppocr_server_v2.0_rec/train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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]}]
test_tipc/configs/ch_ppocr_server_v2.0_rec/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/det_mv3_db_v2_0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/det_mv3_east_v2.0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/det_mv3_pse_v2.0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/det_r18_vd_db_v2_0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/det_r50_db_v2.0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/det_r50_vd_east_v2_0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/det_r50_vd_pse_v2_0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/det_r50_vd_sast_icdar15_v2.0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/det_r50_vd_sast_totaltext_v2.0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/en_server_pgnetA/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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
......
test_tipc/configs/en_table_structure/table_mv3.yml 0 → 100755
浏览文件 @ 5016894e
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
test_tipc/configs/en_table_structure/train_infer_python.txt 0 → 100644
浏览文件 @ 5016894e
===========================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]}]
test_tipc/configs/en_table_structure/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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]}]
test_tipc/configs/en_table_structure_KL/model_linux_gpu_normal_normal_infer_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================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
test_tipc/configs/en_table_structure_PACT/train_infer_python.txt 0 → 100644
浏览文件 @ 5016894e
===========================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]}]
test_tipc/configs/en_table_structure_PACT/train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt 0 → 100644
浏览文件 @ 5016894e
===========================train_params===========================
model_name:en_table_structure_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=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]}]
test_tipc/configs/rec_mtb_nrtr/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/rec_mv3_none_none_ctc_v2.0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/rec_mv3_tps_bilstm_att_v2.0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/rec_r31_sar/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/rec_r34_vd_tps_bilstm_att_v2.0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/configs/rec_r50_fpn_vd_none_srn/train_infer_python.txt
浏览文件 @ 5016894e
......@@ -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/
......
test_tipc/docs/test_serving.md
浏览文件 @ 5016894e
# 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
......
test_tipc/docs/test_train_fleet_inference_python.md 0 → 100644
浏览文件 @ 5016894e
# 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。
test_tipc/prepare.sh
浏览文件 @ 5016894e
......@@ -44,37 +44,46 @@ 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} =~ "PPOCRv3_det" ]];then
# TODO check model link
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_PPOCRv3_rec" ] || [ ${model_name} == "ch_PPOCRv3_rec_PACT" ]; then
# TODO check model link
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
......@@ -89,8 +98,9 @@ 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 ./train_data && tar xf total_text_lite.tar && ln -s total_text_lite total_text && 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
......@@ -114,15 +124,19 @@ 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_PPOCRv3_det" ]; then
# TODO check model link
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
......@@ -137,24 +151,41 @@ 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_PPOCRv3_det" ]; then
# TODO check model link
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
......@@ -199,31 +230,31 @@ elif [ ${MODE} = "whole_infer" ];then
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_PPOCRv2_det" ]]; then
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} =~ "ch_PPOCRv3_det" ]]; then
# TODO check model link
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} =~ "PPOCRv2_ocr_rec" ]]; then
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} =~ "PPOCRv3_ocr_rec" ]]; then
# TODO check model link
if [[ ${model_name} =~ "ch_PP-OCRv3_rec" ]]; then
eval_model_name="ch_PP-OCRv3_rec_infer"
echo $(pwd)
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_quant_infer.tar --no-check-certificate
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
# cd ./inference && tar xf ${eval_model_name}.tar && tar xf ch_PP-OCRv3_rec_slim_quant_infer.tar && cd ../
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
......@@ -314,6 +345,12 @@ 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
......@@ -324,28 +361,29 @@ 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} = "PPOCRv3_ocr_rec_kl" ]; then
# TODO check model link
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 && cd ../
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} = "PPOCRv2_ocr_det_kl" ]; then
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} = "PPOCRv3_ocr_det_kl" ]; then
# TODO check model link
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 ../
......@@ -357,31 +395,113 @@ if [ ${MODE} = "klquant_whole_infer" ]; then
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
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} = "ocr_det_v3" ]; then
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} = "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} = "ocr_rec_v3" ]; then
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_ppocr_mobile_v2.0_rec" ]; then
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/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
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} = "ocr_system_v3" ]; then
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
......@@ -395,36 +515,84 @@ 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
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tar
${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_PP-OCRv3_det_infer.tar && tar xf ch_PP-OCRv3_rec_infer.tar && tar xf rec_inference.tar && cd ../
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 ../
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 paddle2onnx
${python_name} -m pip install onnxruntime==1.4.0
${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
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_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 [[ ${model_name} =~ "ch_PP-OCRv3" ]]; then
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar --no-check-certificate
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tar --no-check-certificate
cd ./inference && tar xf ch_PP-OCRv3_det_infer.tar && tar xf ch_PP-OCRv3_rec_infer.tar && cd ../
fi
# wget data
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar
cd ./inference && tar xf ch_ppocr_mobile_v2.0_det_infer.tar && tar xf ch_ppocr_mobile_v2.0_rec_infer.tar && tar xf ch_ppocr_server_v2.0_rec_infer.tar && tar xf ch_ppocr_server_v2.0_det_infer.tar && tar xf ch_det_data_50.tar && tar xf rec_inference.tar && cd ../
cd ./inference && tar xf ch_det_data_50.tar && tar xf rec_inference.tar && cd ../
fi
test_tipc/readme.md
浏览文件 @ 5016894e
......@@ -138,6 +138,7 @@ bash test_tipc/test_train_inference_python.sh ./test_tipc/configs/ch_ppocr_mobil
## 4. 开始测试
各功能测试中涉及混合精度、裁剪、量化等训练相关，及mkldnn、Tensorrt等多种预测相关参数配置，请点击下方相应链接了解更多细节和使用教程：
- [test_train_inference_python 使用](docs/test_train_inference_python.md) ：测试基于Python的模型训练、评估、推理等基本功能，包括裁剪、量化、蒸馏。
- [test_train_fleet_inference_python 使用](./docs/test_train_fleet_inference_python.md)：测试基于Python的多机多卡训练与推理等基本功能。
- [test_inference_cpp 使用](docs/test_inference_cpp.md)：测试基于C++的模型推理。
- [test_serving 使用](docs/test_serving.md)：测试基于Paddle Serving的服务化部署功能。
- [test_lite_arm_cpp 使用](docs/test_lite_arm_cpp.md)：测试基于Paddle-Lite的ARM CPU端c++预测部署功能。
......
test_tipc/test_inference_cpp.sh
浏览文件 @ 5016894e
......@@ -43,7 +43,7 @@ cpp_cls_value=$(func_parser_value "${lines[18]}")
cpp_use_angle_cls_key=$(func_parser_key "${lines[19]}")
cpp_use_angle_cls_value=$(func_parser_value "${lines[19]}")
LOG_PATH="./test_tipc/output/cpp_infer/${model_name}"
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
......@@ -178,6 +178,22 @@ if [ ${use_opencv} = "True" ]; then
else
OPENCV_DIR=''
fi
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
......
test_tipc/test_inference_python.sh
浏览文件 @ 5016894e
......@@ -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
......@@ -119,7 +119,7 @@ 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
......@@ -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=${infer_model}
fi
......
test_tipc/test_paddle2onnx.sh
浏览文件 @ 5016894e
......@@ -11,7 +11,7 @@ python=$(func_parser_value "${lines[2]}")
# parser params
dataline=$(awk 'NR==1, NR==14{print}' $FILENAME)
dataline=$(awk 'NR==1, NR==17{print}' $FILENAME)
IFS=$'\n'
lines=(${dataline})
......@@ -19,27 +19,32 @@ lines=(${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_list=$(func_parser_value "${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/paddle2onnx/${model_name}"
LOG_PATH="./test_tipc/output/${model_name}/paddle2onnx"
mkdir -p ${LOG_PATH}
status_log="${LOG_PATH}/results_paddle2onnx.log"
......@@ -49,38 +54,95 @@ function func_paddle2onnx(){
_script=$1
# paddle2onnx
set_dirname=$(func_set_params "${infer_model_dir_key}" "${infer_model_dir_value}")
if [ ${model_name} = "ch_PP-OCRv2" ] || [ ${model_name} = "ch_PP-OCRv3" ] || [ ${model_name} = "ch_ppocr_mobile_v2.0" ] || [ ${model_name} = "ch_ppocr_server_v2.0" ]; then
# trans det
set_dirname=$(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 "${save_file_key}" "${save_file_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_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}"
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_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}" "${status_log}" "${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}" "${enable_onnx_checker_value}")
trans_model_cmd="${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}" "${status_log}" "${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}" "${enable_onnx_checker_value}")
trans_model_cmd="${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}" "${status_log}" "${model_name}"
fi
# python inference
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}" "${use_gpu}")
set_model_dir=$(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 "
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}"
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}" "${use_gpu}")
set_model_dir=$(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 "
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}"
status_check $last_status "${infer_model_cmd}" "${status_log}" "${model_name}"
else
echo "Does not support hardware other than CPU and GPU Currently!"
fi
......
test_tipc/test_serving.sh 已删除 100644 → 0
浏览文件 @ 16bafede
#!/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 "${lines[17]}")
LOG_PATH="$(pwd)/test_tipc/output/pdserving/${model_name}"
mkdir -p ${LOG_PATH}
status_log="${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
if [ ${model_name} = "ocr_det_mobile" ]; then
web_service_cpp_cmd="${python_list[0]} -m paddle_serving_server.serve --model ppocr_det_mobile_2.0_serving/ ppocr_rec_mobile_2.0_serving/ --op GeneralDetectionOp GeneralInferOp --port 9293"
elif [ ${model_name} = "ocr_det_v3" ] || [ ${model_name} = "ocr_rec_v3" ]; then
web_service_cpp_cmd="${python_list[0]} -m paddle_serving_server.serve --model ppocr_det_v3_serving/ ppocr_rec_v3_serving/ --op GeneralDetectionOp GeneralInferOp --port 9293"
fi
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"
if [ ${model_name} = "ocr_det_mobile" ]; then
pipeline_cmd="${python_list[0]} ocr_cpp_client.py ppocr_det_mobile_2.0_client/ ppocr_rec_mobile_2.0_client/"
elif [ ${model_name} = "ocr_det_v3" ] || [ ${model_name} = "ocr_rec_v3" ]; then
pipeline_cmd="${python_list[0]} ocr_cpp_client.py ppocr_det_v3_client/ ppocr_rec_v3_client/"
fi
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
if [ ${model_name} = "ocr_det_mobile" ]; then
web_service_cpp_cmd="${python_list[0]} -m paddle_serving_server.serve --model ppocr_det_mobile_2.0_serving/ ppocr_rec_mobile_2.0_serving/ --op GeneralDetectionOp GeneralInferOp --port 9293 --gpu_id=0"
elif [ ${model_name} = "ocr_det_v3" ] || [ ${model_name} = "ocr_rec_v3" ]; then
web_service_cpp_cmd="${python_list[0]} -m paddle_serving_server.serve --model ppocr_det_v3_serving/ ppocr_rec_v3_serving/ --op GeneralDetectionOp GeneralInferOp --port 9293 --gpu_id=0"
fi
eval $web_service_cpp_cmd
sleep 2s
_save_log_path="${LOG_PATH}/server_infer_cpp_cpu_pipeline_usemkldnn_False_threads_4_batchsize_1.log"
if [ ${model_name} = "ocr_det_mobile" ]; then
pipeline_cmd="${python_list[0]} ocr_cpp_client.py ppocr_det_mobile_2.0_client/ ppocr_rec_mobile_2.0_client/"
elif [ ${model_name} = "ocr_det_v3" ] || [ ${model_name} = "ocr_rec_v3" ]; then
pipeline_cmd="${python_list[0]} ocr_cpp_client.py ppocr_det_v3_client/ ppocr_rec_v3_client/"
fi
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} = "gpu" ]; 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} ${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}"
test_tipc/test_serving_infer_cpp.sh 0 → 100644
浏览文件 @ 5016894e
#!/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_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
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_model_cmd="${python_list[0]} ${trans_model_py} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_serving_server} ${set_serving_client}"
eval $trans_model_cmd
last_status=${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 [ ${gpu_id} = "null" ]; then
web_service_cpp_cmd="${python_list[0]} ${web_service_py} --model ${det_server_value} ${rec_server_value} ${op_key} ${op_value} ${port_key} ${port_value} > serving_log_cpu.log &"
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="${LOG_PATH}/server_infer_cpp_cpu.log"
cpp_client_cmd="${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]}
status_check $last_status "${cpp_client_cmd}" "${status_log}" "${model_name}"
ps ux | grep -i ${port_value} | awk '{print $2}' | xargs kill -s 9
else
web_service_cpp_cmd="${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} > serving_log_gpu.log &"
eval $web_service_cpp_cmd
sleep 5s
_save_log_path="${LOG_PATH}/server_infer_cpp_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}"
test_tipc/test_serving_infer_python.sh 0 → 100644
浏览文件 @ 5016894e
#!/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_model_cmd="${python_list[0]} ${trans_model_py} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_serving_server} ${set_serving_client}"
eval $trans_model_cmd
# trans rec
set_dirname=$(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_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
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_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
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_model_cmd="${python_list[0]} ${trans_model_py} ${set_dirname} ${set_model_filename} ${set_params_filename} ${set_serving_server} ${set_serving_client}"
eval $trans_model_cmd
fi
set_image_dir=$(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}")
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} &"
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} &"
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} &"
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}/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}" "${status_log}" "${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
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} &"
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} &"
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} &"
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}/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}" "${status_log}" "${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}"
test_tipc/test_train_inference_python.sh
浏览文件 @ 5016894e
......@@ -142,9 +142,9 @@ 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 +169,7 @@ 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 +200,7 @@ 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
......@@ -240,7 +240,7 @@ if [ ${MODE} = "whole_infer" ] || [ ${MODE} = "klquant_whole_infer" ]; then
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=${infer_model}
fi
......@@ -315,7 +315,9 @@ else
set_batchsize=$(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_${autocast}"
nodes=1
else
......@@ -330,14 +332,14 @@ else
set_save_model=$(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}")
......@@ -347,7 +349,7 @@ else
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 $eval_cmd
status_check $? "${eval_cmd}" "${status_log}"
status_check $? "${eval_cmd}" "${status_log}" "${model_name}"
fi
# run export model
if [ ${run_export} != "null" ]; then
......@@ -357,7 +359,7 @@ else
set_save_infer_key=$(func_set_params "${save_infer_key}" "${save_infer_path}")
export_cmd="${python} ${run_export} ${set_export_weight} ${set_save_infer_key}"
eval $export_cmd
status_check $? "${export_cmd}" "${status_log}"
status_check $? "${export_cmd}" "${status_log}" "${model_name}"
#run inference
eval $env
......
tools/end2end/convert_ppocr_label.py
浏览文件 @ 5016894e
......@@ -85,10 +85,16 @@ def convert_label(label_dir, mode="gt", save_dir="./save_results/"):
print("The convert label saved in {}".format(save_dir))
if __name__ == "__main__":
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
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/")
if __name__ == "__main__":
args = parse_args()
convert_label(args.label_path, args.mode, args.save_folder)
tools/end2end/readme.md
浏览文件 @ 5016894e
......@@ -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。
```
ppocr_label_gt = "gt_label.txt"
convert_label(ppocr_label_gt, "gt", "./save_gt_label/")
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 = "./ch_PP-OCRv2_results/system_results.txt"
convert_label(ppocr_label_gt_en, "pred", "./save_PPOCRV2_infer/")
```
python3 tools/end2end/convert_ppocr_label.py --mode=gt --label_path=path/to/label_txt --save_folder=save_gt_label
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：
......
tools/export_model.py
浏览文件 @ 5016894e
......@@ -76,7 +76,7 @@ def export_single_model(model, arch_config, save_path, logger, quanter=None):
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":
......
tools/infer/utility.py
浏览文件 @ 5016894e
......@@ -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":
......@@ -286,6 +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"):
......
tools/program.py
浏览文件 @ 5016894e
......@@ -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
......@@ -547,7 +552,7 @@ 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
......@@ -562,11 +567,13 @@ def preprocess(is_train=False):
'SEED', 'SDMGR', 'LayoutXLM', 'LayoutLM', 'PREN', 'FCE', 'SVTR'
]
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
......
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或