search

xiaofengShi / CHINESE-OCR

[python3.6] 运用tf实现自然场景文字检测,keras/pytorch实现ctpn+crnn+ctc实现不定长场景文字OCR识别
2.92k stars 964 forks source link

运行trainbatch报错：utf8 code canlt decode #140

Open mrlihellohorld opened 4 years ago

mrlihellohorld commented 4 years ago

制作好了自己的数据集开始训练的时候，运行trainbatch.py的时候一直报错：UnicodeDecodeError:'utf8'code can;t decode byte 0xe6 in position 0: unixpected end of data。能给我一点思路吗，谢谢