Tao Luo
Tao Luo
@lyn0102 能具体说下么？或者po图。
而且好像需要手动改mysql ip，不如直接来个compose来的方便。
@oyxhust so,you mean you just want to converge faster?
#17 the ctc loss comes to negative is common,you can search it on google. ^.^ As above your error info mentioned,you need covert your convert your tensor to cpu,the tensor...
not correct... ---Original--- From: "Xiangpeng Li"
@lixiangpengcs replace the original `ctc_loss` with [torch-baidu-ctc](https://pypi.org/project/torch-baidu-ctc/). because of the bug of ctc loss....
if the error still comes out ,i think you need checkout your gt
@lixiangpengcs how many pcs of your dataset?
i modified the recognition part code.you can modify for your application. ------------------ 原始邮件 ------------------ 发件人: "Xiangpeng Li"; 发送时间: 2019年9月18日(星期三) 晚上8:30 收件人: "novioleo/FOTS"; 抄送: "北国枫叶。"; "Mention"; 主题: Re: [novioleo/FOTS] Confiusion about...
icdar2015 usually cause some problems,you can try your own dataset ---Original--- From: "Xiangpeng Li"