always training the first 100 batch

Belval / CRNN

A TensorFlow implementation of https://github.com/bgshih/crnn

MIT License

297 stars 101 forks source link

Open lulupanPPY opened 4 years ago

lulupanPPY commented 4 years ago

plz take a look at line CRNN.py line 308 and 309, i think you train only the first 100 batch