issues
search
Hironsan
/
anago
Bidirectional LSTM-CRF and ELMo for Named-Entity Recognition, Part-of-Speech Tagging and so on.
https://anago.herokuapp.com/
MIT License
1.48k
stars
368
forks
source link
Update trainer.py and wrapper.py
#22
Closed
sehoi
closed
6 years ago
sehoi
commented
6 years ago
to avoid crash at training, add "verbose" parameter at train methods
add return statement for check train result
increase default batch size to 1024 for use more CPU at training