Open Harryoung opened 1 year ago
#
configs/rec/PP-OCRv3/en_PP-OCRv3_rec.yml
,在其中添加character_dict_path
字段指向自定义字典的路径,并将use_space_char
字段设置为True
以支持识别空格类别。https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.7/doc/doc_ch/FAQ.md, https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.7/doc/doc_ch/recognition.md
My answer may not be accurate. I suggest you refer to relevant sources to make a comprehensive judgment. To invoke me for more questions, add a comment using following command: @llm-issue-bot \<QUESTION>
我们提供了issue bot来帮助你解答问题,你是否想要它来解答(请填写yes/no)?/We provide issue bot to solve issues, do you want it? (Please write yes/no):yes
请在这里描述你的问题/Please describe your issue here:
您好,很感谢您的工作,想咨询两个问题 1.可以不使用dict,输入的文本标签可以使用已经转化过后的数字做训练吗? 2.使用dict,空格需要增加在字典里吗,训练需要修改那些代码文件呢?