forgetest5
/
PaddleOCR
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Update model conversion instructions

This commit is contained in:
WenmuZhou 2020-12-10 00:26:19 +08:00
parent 0ff2aef299
commit 0a28221d76
2 changed files with 37 additions and 34 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
doc/doc_ch/inference.md
View File

@ -45,7 +45,7 @@ wget -P ./ch_lite/ {link} && tar xf ./ch_lite/{file} -C ./ch_lite/
```
上述模型是以MobileNetV3为backbone训练的DB算法，将训练好的模型转换成inference模型只需要运行如下命令：
```
# -c 后面设置训练算法的yml配置文件，需将待转换的训练模型地址写入配置文件里的Global.checkpoints字段下, 不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -c 后面设置训练算法的yml配置文件，需设置 `Global.load_static_weights=False`, 并将待转换的训练模型地址写入配置文件里的 `Global.pretrained_model` 字段下，不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -o 后面设置转换的模型将保存的地址。
python3 tools/export_model.py -c configs/det/det_mv3_db_v1.1.yml -o ./inference/det_db/
@ -54,9 +54,9 @@ python3 tools/export_model.py -c configs/det/det_mv3_db_v1.1.yml -o ./inference/
转换成功后，在模型保存目录下有三个文件：
```
inference/det_db/
├── det.pdiparams # 检测inference模型的参数文件
├── det.pdiparams.info # 检测inference模型的参数信息，可忽略
└── det.pdmodel # 检测inference模型的program文件
├── inference.pdiparams # 检测inference模型的参数文件
├── inference.pdiparams.info # 检测inference模型的参数信息，可忽略
└── inference.pdmodel # 检测inference模型的program文件
```
<a name="识别模型转inference模型"></a>
@ -69,7 +69,7 @@ wget -P ./ch_lite/ {link} && tar xf ./ch_lite/{file} -C ./ch_lite/
识别模型转inference模型与检测的方式相同，如下：
```
# -c 后面设置训练算法的yml配置文件，需将待转换的训练模型地址写入配置文件里的Global.checkpoints字段下,不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -c 后面设置训练算法的yml配置文件，需设置 `Global.load_static_weights=False`, 并将待转换的训练模型地址写入配置文件里的 `Global.pretrained_model` 字段下，不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -o 后面设置转换的模型将保存的地址。
python3 tools/export_model.py -c configs/rec/ch_ppocr_v1.1/rec_chinese_lite_train_v1.1.yml -o ./inference/rec_crnn/
```
@ -79,9 +79,9 @@ python3 tools/export_model.py -c configs/rec/ch_ppocr_v1.1/rec_chinese_lite_trai
转换成功后，在目录下有三个文件：
```
/inference/rec_crnn/
├── rec.pdiparams # 识别inference模型的参数文件
├── rec.pdiparams.info # 识别inference模型的参数信息，可忽略
└── rec.pdmodel # 识别inference模型的program文件
├── inference.pdiparams # 识别inference模型的参数文件
├── inference.pdiparams.info # 识别inference模型的参数信息，可忽略
└── inference.pdmodel # 识别inference模型的program文件
```
<a name="方向分类模型转inference模型"></a>
@ -94,7 +94,7 @@ wget -P ./ch_lite/ {link} && tar xf ./ch_lite/{file} -C ./ch_lite/
方向分类模型转inference模型与检测的方式相同，如下：
```
# -c 后面设置训练算法的yml配置文件，需将待转换的训练模型地址写入配置文件里的Global.checkpoints字段下,不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -c 后面设置训练算法的yml配置文件，需设置 `Global.load_static_weights=False`, 并将待转换的训练模型地址写入配置文件里的 `Global.pretrained_model` 字段下，不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -o 后面设置转换的模型将保存的地址。
python3 tools/export_model.py -c configs/cls/cls_mv3.yml -o ./inference/cls/
@ -103,9 +103,9 @@ python3 tools/export_model.py -c configs/cls/cls_mv3.yml -o ./inference/cls/
转换成功后，在目录下有三个文件：
```
/inference/cls/
├── cls.pdiparams # 分类inference模型的参数文件
├── cls.pdiparams.info # 分类inference模型的参数信息，可忽略
└── cls.pdmodel # 分类inference模型的program文件
├── inference.pdiparams # 分类inference模型的参数文件
├── inference.pdiparams.info # 分类inference模型的参数信息，可忽略
└── inference.pdmodel # 分类inference模型的program文件
```
<a name="文本检测模型推理"></a>
@ -145,7 +145,7 @@ python3 tools/infer/predict_det.py --image_dir="./doc/imgs/2.jpg" --det_model_di
首先将DB文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd骨干网络，在ICDAR2015英文数据集训练的模型为例（[模型下载地址](link))，可以使用如下命令进行转换：
```
# -c 后面设置训练算法的yml配置文件，需将待转换的训练模型地址写入配置文件里的Global.checkpoints字段下,不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -c 后面设置训练算法的yml配置文件，需设置 `Global.load_static_weights=False`, 并将待转换的训练模型地址写入配置文件里的 `Global.pretrained_model` 字段下，不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -o 后面设置转换的模型将保存的地址。
python3 tools/export_model.py -c configs/det/det_r50_vd_db.yml -o "./inference/det_db"
@ -169,7 +169,7 @@ python3 tools/infer/predict_det.py --image_dir="./doc/imgs_en/img_10.jpg" --det_
首先将EAST文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd骨干网络，在ICDAR2015英文数据集训练的模型为例（[模型下载地址](link))，可以使用如下命令进行转换：
```
# -c 后面设置训练算法的yml配置文件，需将待转换的训练模型地址写入配置文件里的Global.checkpoints字段下,不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -c 后面设置训练算法的yml配置文件，需设置 `Global.load_static_weights=False`, 并将待转换的训练模型地址写入配置文件里的 `Global.pretrained_model` 字段下，不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -o 后面设置转换的模型将保存的地址。
python3 tools/export_model.py -c configs/det/det_r50_vd_east.yml -o Global.checkpoints="./models/det_r50_vd_east/best_accuracy" Global.save_inference_dir="./inference/det_east"
@ -192,7 +192,7 @@ python3 tools/infer/predict_det.py --det_algorithm="EAST" --image_dir="./doc/img
#### (1). 四边形文本检测模型（ICDAR2015）
首先将SAST文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd骨干网络，在ICDAR2015英文数据集训练的模型为例([模型下载地址](link))，可以使用如下命令进行转换：
```
# -c 后面设置训练算法的yml配置文件，需将待转换的训练模型地址写入配置文件里的Global.checkpoints字段下,不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -c 后面设置训练算法的yml配置文件，需设置 `Global.load_static_weights=False`, 并将待转换的训练模型地址写入配置文件里的 `Global.pretrained_model` 字段下，不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -o 后面设置转换的模型将保存的地址。
python3 tools/export_model.py -c configs/det/det_r50_vd_sast_icdar15.yml -o "./inference/det_sast_ic15"
@ -209,7 +209,7 @@ python3 tools/infer/predict_det.py --det_algorithm="SAST" --image_dir="./doc/img
首先将SAST文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd骨干网络，在Total-Text英文数据集训练的模型为例（[模型下载地址](link))，可以使用如下命令进行转换：
```
# -c 后面设置训练算法的yml配置文件，需将待转换的训练模型地址写入配置文件里的Global.checkpoints字段下,不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -c 后面设置训练算法的yml配置文件，需设置 `Global.load_static_weights=False`, 并将待转换的训练模型地址写入配置文件里的 `Global.pretrained_model` 字段下，不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -o 后面设置转换的模型将保存的地址。
python3 tools/export_model.py -c configs/det/det_r50_vd_sast_totaltext.yml -o "./inference/det_sast_tt"
@ -257,7 +257,7 @@ Predicts of ./doc/imgs_words/ch/word_4.jpg:['实力活力', 0.89552695]
的模型为例（[模型下载地址](link))，可以使用如下命令进行转换：
```
# -c 后面设置训练算法的yml配置文件，需将待转换的训练模型地址写入配置文件里的Global.checkpoints字段下,不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -c 后面设置训练算法的yml配置文件，需设置 `Global.load_static_weights=False`, 并将待转换的训练模型地址写入配置文件里的 `Global.pretrained_model` 字段下，不用添加文件后缀.pdmodel，.pdopt或.pdparams。
# -o 后面设置转换的模型将保存的地址。
python3 tools/export_model.py -c configs/rec/rec_r34_vd_tps_bilstm_ctc.yml -o "./inference/starnet"

35
doc/doc_en/inference_en.md
View File

@ -48,7 +48,7 @@ wget -P ./ch_lite/ {link} && tar xf ./ch_lite/{file} -C ./ch_lite/
The above model is a DB algorithm trained with MobileNetV3 as the backbone. To convert the trained model into an inference model, just run the following command:
```
# -c Set the yml configuration file of the training algorithm, you need to write the path of the training model to be converted into the Global.checkpoints parameter in the configuration file, without adding the file suffixes .pdmodel, .pdopt or .pdparams.
-c Set the yml configuration file of the algorithm, you need to set `Global.load_static_weights=False`, and write the path of the training model to be converted under the `Global.pretrained_model` parameter in the configuration file, without adding the file suffix .pdmodel, .pdopt or .pdparams.
# -o Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/det/det_mv3_db_v1.1.yml -o ./inference/det_db/
@ -58,9 +58,9 @@ When converting to an inference model, the configuration file used is the same a
After the conversion is successful, there are three files in the model save directory:
```
inference/det_db/
├── det.pdiparams # The parameter file of detection inference model
├── det.pdiparams.info # The parameter information of detection inference model, which can be ignored
└── det.pdmodel # The program file of detection inference model
├── inference.pdiparams # The parameter file of detection inference model
├── inference.pdiparams.info # The parameter information of detection inference model, which can be ignored
└── inference.pdmodel # The program file of detection inference model
```
<a name="Convert_recognition_model"></a>
@ -73,7 +73,7 @@ wget -P ./ch_lite/ {link} && tar xf ./ch_lite/{file} -C ./ch_lite/
The recognition model is converted to the inference model in the same way as the detection, as follows:
```
# -c Set the yml configuration file of the training algorithm, you need to write the path of the training model to be converted into the Global.checkpoints parameter in the configuration file, without adding the file suffixes .pdmodel, .pdopt or .pdparams.
-c Set the yml configuration file of the algorithm, you need to set `Global.load_static_weights=False`, and write the path of the training model to be converted under the `Global.pretrained_model` parameter in the configuration file, without adding the file suffix .pdmodel, .pdopt or .pdparams.
# -o Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/cls/cls_mv3.yml -o ./inference/cls/
@ -84,9 +84,9 @@ If you have a model trained on your own dataset with a different dictionary file
After the conversion is successful, there are three files in the model save directory:
```
inference/det_db/
├── rec.pdiparams # The parameter file of recognition inference model
├── rec.pdiparams.info # The parameter information of recognition inference model, which can be ignored
└── rec.pdmodel # The program file of recognition model
├── inference.pdiparams # The parameter file of recognition inference model
├── inference.pdiparams.info # The parameter information of recognition inference model, which can be ignored
└── inference.pdmodel # The program file of recognition model
```
<a name="Convert_angle_class_model"></a>
@ -99,7 +99,7 @@ wget -P ./ch_lite/ {link} && tar xf ./ch_lite/{file} -C ./ch_lite/
The angle classification model is converted to the inference model in the same way as the detection, as follows:
```
# -c Set the yml configuration file of the training algorithm, you need to write the path of the training model to be converted into the Global.checkpoints parameter in the configuration file, without adding the file suffixes .pdmodel, .pdopt or .pdparams.
-c Set the yml configuration file of the algorithm, you need to set `Global.load_static_weights=False`, and write the path of the training model to be converted under the `Global.pretrained_model` parameter in the configuration file, without adding the file suffix .pdmodel, .pdopt or .pdparams.
# -o Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/cls/cls_mv3.yml -o ./inference/cls/
@ -108,9 +108,9 @@ python3 tools/export_model.py -c configs/cls/cls_mv3.yml -o ./inference/cls/
After the conversion is successful, there are two files in the directory:
```
inference/det_db/
├── rec.pdiparams # The parameter file of angle class inference model
├── rec.pdiparams.info # The parameter information of angle class inference model, which can be ignored
└── rec.pdmodel # The program file of angle class model
├── inference.pdiparams # The parameter file of angle class inference model
├── inference.pdiparams.info # The parameter information of angle class inference model, which can be ignored
└── inference.pdmodel # The program file of angle class model
```
@ -152,7 +152,7 @@ python3 tools/infer/predict_det.py --image_dir="./doc/imgs/2.jpg" --det_model_di
First, convert the model saved in the DB text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the ICDAR2015 English dataset as an example ([model download link](link)), you can use the following command to convert:
```
# -c Set the yml configuration file of the training algorithm, you need to write the path of the training model to be converted into the Global.checkpoints parameter in the configuration file, without adding the file suffixes .pdmodel, .pdopt or .pdparams.
-c Set the yml configuration file of the algorithm, you need to set `Global.load_static_weights=False`, and write the path of the training model to be converted under the `Global.pretrained_model` parameter in the configuration file, without adding the file suffix .pdmodel, .pdopt or .pdparams.
# -o Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/det/det_r50_vd_db.yml -o "./inference/det_db"
@ -176,7 +176,7 @@ The visualized text detection results are saved to the `./inference_results` fol
First, convert the model saved in the EAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the ICDAR2015 English dataset as an example ([model download link](link)), you can use the following command to convert:
```
# -c Set the yml configuration file of the training algorithm, you need to write the path of the training model to be converted into the Global.checkpoints parameter in the configuration file, without adding the file suffixes .pdmodel, .pdopt or .pdparams.
-c Set the yml configuration file of the algorithm, you need to set `Global.load_static_weights=False`, and write the path of the training model to be converted under the `Global.pretrained_model` parameter in the configuration file, without adding the file suffix .pdmodel, .pdopt or .pdparams.
# -o Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/det/det_r50_vd_east.yml -o Global.checkpoints="./models/det_r50_vd_east/best_accuracy" Global.save_inference_dir="./inference/det_east"
@ -200,7 +200,7 @@ The visualized text detection results are saved to the `./inference_results` fol
First, convert the model saved in the SAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the ICDAR2015 English dataset as an example ([model download link](link)), you can use the following command to convert:
```
# -c Set the yml configuration file of the training algorithm, you need to write the path of the training model to be converted into the Global.checkpoints parameter in the configuration file, without adding the file suffixes .pdmodel, .pdopt or .pdparams.
-c Set the yml configuration file of the algorithm, you need to set `Global.load_static_weights=False`, and write the path of the training model to be converted under the `Global.pretrained_model` parameter in the configuration file, without adding the file suffix .pdmodel, .pdopt or .pdparams.
# -o Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/det/det_r50_vd_sast_icdar15.yml -o "./inference/det_sast_ic15"
@ -220,6 +220,9 @@ The visualized text detection results are saved to the `./inference_results` fol
First, convert the model saved in the SAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the Total-Text English dataset as an example ([model download link](https://paddleocr.bj.bcebos.com/SAST/sast_r50_vd_total_text.tar)), you can use the following command to convert:
```
-c Set the yml configuration file of the algorithm, you need to set `Global.load_static_weights=False`, and write the path of the training model to be converted under the `Global.pretrained_model` parameter in the configuration file, without adding the file suffix .pdmodel, .pdopt or .pdparams.
# -o Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/det/det_r50_vd_sast_totaltext.yml -o Global.checkpoints="./models/sast_r50_vd_total_text/best_accuracy" Global.save_inference_dir="./inference/det_sast_tt"
```
@ -265,7 +268,7 @@ Taking STAR-Net as an example, we introduce the recognition model inference base
First, convert the model saved in the STAR-Net text recognition training process into an inference model. Taking the model based on Resnet34_vd backbone network, using MJSynth and SynthText (two English text recognition synthetic datasets) for training, as an example ([model download address](link)). It can be converted as follow:
```
# -c Set the yml configuration file of the training algorithm, you need to write the path of the training model to be converted into the Global.checkpoints parameter in the configuration file, without adding the file suffixes .pdmodel, .pdopt or .pdparams.
-c Set the yml configuration file of the algorithm, you need to set `Global.load_static_weights=False`, and write the path of the training model to be converted under the `Global.pretrained_model` parameter in the configuration file, without adding the file suffix .pdmodel, .pdopt or .pdparams.
# -o Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/rec/rec_r34_vd_tps_bilstm_ctc.yml -o "./inference/starnet"