1 Release published by 1 person

• v2.8.0

  published Jul 4, 2024

7 Pull requests merged by 3 people

• unifying data types in the SLAHead

  #13276 merged Jul 6, 2024

• add url in pyproject, and update version number

  #13274 merged Jul 6, 2024

• fix bug when layout_predictor is None

  #13279 merged Jul 6, 2024

• fix: table recognition content is not escaped properly

  #13277 merged Jul 6, 2024

• update ReadMe

  #13263 merged Jul 4, 2024

• Updated Recognition Competition Model Link

  #13259 merged Jul 4, 2024

• The model link for tabe rec compti

  #13254 merged Jul 3, 2024

4 Pull requests opened by 3 people

• ..

  #13226 opened Jul 1, 2024

• Update expired link in quickstart.md

  #13253 opened Jul 3, 2024

• optimize func: get_infer_gpuid

  #13275 opened Jul 6, 2024

• [cherry-pick] add project url and fix a bug

  #13281 opened Jul 6, 2024

15 Issues closed by 4 people

• 表格识别 训练时候，进行eval 时候遇到bug

  #13268 closed Jul 6, 2024

• 表格识别内容没有正常转义 '<' '>'

  #12509 closed Jul 6, 2024

• AttributeError: 'bool' object has no attribute 'sum' when `use_gpu: False`

  #12132 closed Jul 6, 2024

• PaddleOCR returning only the first page when performing ocr on a PDF

  #13094 closed Jul 6, 2024

• 使用PP-OCRv3做文字检测微调，在T4上跑训练无法训练，报错返回-9

  #13100 closed Jul 3, 2024

• 文本被检测框拼接了，这种需要自己训练模型么？

  #13245 closed Jul 3, 2024

• ModuleNotFoundError: No module named `common` even when common is installed

  #13246 closed Jul 2, 2024

• 请问训练det_server模型时候可以使用混合精度？

  #13242 closed Jul 2, 2024

• 版面分析正确率上不去

  #13231 closed Jul 2, 2024

• svtr+stn，inference和onnx模型推理精度差异很大

  #13230 closed Jul 2, 2024

• ch_PP-OCRv4_server检测不出灰烬的烬字

  #13236 closed Jul 2, 2024

• 有办法检测字体缺失吗

  #13235 closed Jul 1, 2024

• picodet_lcnet_x1_0_layout.yml更换模型#paddleocr更换模型

  #13220 closed Jun 30, 2024

• ppocr/data/simple_dataset.py第63行有个代码bug

  #13219 closed Jun 29, 2024

• 使用ch_PP-OCRv4_rec.yml训练模型报错AttributeError: 'SimpleDataSet' object has no attribute 'ds_width'

  #13217 closed Jun 29, 2024

10 Issues opened by 8 people

• paddle version compatible

  #13280 opened Jul 6, 2024

• 关于执行PP-Structure文档的命令都提示报错，paddleocr --image_dir=ppstructure/docs/table/1.png --type=structure --image_orientation=true

  #13278 opened Jul 6, 2024

• PGNet model downloaded does not perform as good as reported.

  #13271 opened Jul 5, 2024

• 关于使用ppstructure版面分析模型3个模型（都是210M大小的模型：）报错，3个没有达到满意的结果

  #13270 opened Jul 5, 2024

• PaddleOCR2.7.5 训练到几百回合出现loss>1问题

  #13267 opened Jul 5, 2024

• I can help with Turkish language.

  #13265 opened Jul 4, 2024

• 关于业务代码Java调用服务端hub serving部署的structure_table报错

  #13260 opened Jul 4, 2024

• Low Validation Accuracy for PPOCRv4 if Image resolution is changed

  #13252 opened Jul 3, 2024

• Jetson TX2 NX 使用GPU预测没问题，使用TRT就卡住

  #13234 opened Jul 1, 2024

• 文本检测微调模型 转推理模型后识别效果不一致

  #13228 opened Jul 1, 2024

5 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• About the bug of paddleocr running with distributed.launch

  #13218 commented on Jul 3, 2024 • 5 new comments

• ch_PP-OCRv4_rec_svtr_large.yml训练导出的模型，使用predict_rec.py预测宽度比较大的图片时出现(InvalidArgument) Broadcast dimension mismatch

  #12440 commented on Jul 6, 2024 • 4 new comments

• RE模型训练出错，谁能帮忙分析一下原因

  #13193 commented on Jul 4, 2024 • 1 new comment

• Replace raw pointer by `std::unique_ptr`

  #8647 commented on Jul 5, 2024 • 1 new comment

• 引用PaddleOCR模块后会导致logging的默认logger中添加一项haddler

  #13122 commented on Jul 3, 2024 • 0 new comments