JavieYuan

Results 5 issues of JavieYuan

请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem - 系统环境/System Environment: - 版本号/Version:Paddle: PaddleOCR: 问题相关组件/Related components: - 运行指令/Command Code: - 完整报错/Complete Error Message: ![图片1](https://user-images.githubusercontent.com/84179587/192769612-1b757e59-135c-46b3-8acf-6ee155f70292.png) - 训练结果如上图所示 - 训练集大概100W张中文文本图,6W张英文文本图,字典包含中英文字符为7千左右,训练数据如下图所示:...

请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem - 系统环境/System Environment: - 版本号/Version:Paddle: PaddleOCR: 问题相关组件/Related components: - 运行指令/Command Code: - 完整报错/Complete Error Message: ![que](https://user-images.githubusercontent.com/84179587/194743904-70ceabea-e737-44f6-831c-0474f9172569.png) 请问关于输入图像32*320这部分,说把图像等比例缩放至3,32,320以及不足320部分补0,关于这部分的源码在哪里呀,好像不在数据增强rec_img_aug里面???

请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem - 系统环境/System Environment: - 版本号/Version:Paddle: PaddleOCR: 问题相关组件/Related components: - 运行指令/Command Code: - 完整报错/Complete Error Message: 识别部分-------自己训练结束的模型在 python tools/eval.py 测试后准确率很高 分别87%,98%,但是相同的数据相同配置在执行python...

请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem - 系统环境/System Environment: - 版本号/Version:Paddle: PaddleOCR: 问题相关组件/Related components: - 运行指令/Command Code: - 完整报错/Complete Error Message: 几个月前拉取的代码,最近训练完后使用python3 tools/export_model.py -c configs/rec/multi_language/premodel_aug3new_4mix.yml -o...

请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem - 系统环境/System Environment: - 版本号/Version:Paddle: PaddleOCR: 问题相关组件/Related components: - 运行指令/Command Code: - 完整报错/Complete Error Message: 将utility.py中参数benchmark打开: ![bug4](https://user-images.githubusercontent.com/84179587/201619428-da7a4820-d5f0-468b-8bee-3ee13764b139.png) 然后使用inference推理模型进行predict_rec.py出现错误: ![bug4-1](https://user-images.githubusercontent.com/84179587/201620493-6df896ef-71b8-4b1e-8f10-5e033c6e7277.png)