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

custom optimizer is not used in trainer.py #45

Closed ismaeIfm closed 6 years ago

ismaeIfm commented 6 years ago

If you define a custom optimizer in a training_config to use it with the Trainer class, the optimizer is never used since Adam is hardcoded as an optimizer. https://github.com/Hironsan/anago/blob/master/anago/trainer.py#L38

ismaeIfm commented 6 years ago

Same issue with lr_decay and clip_gradients. I would suggest using only optimizer as parameter the same way keras handles it

Hironsan commented 6 years ago

In anaGo 1.0.0, this problem is solved. Thanks!

Example:

model = anago.Sequence(optimizer='rmsprop')
model.fit(x_train, y_train)