search

Cranial-XIX / Generative-Neural-Parser

generative neural parser
1 stars 0 forks source link

LSTM-minus implementation #2

Open wddabc opened 7 years ago

wddabc commented 7 years ago

https://github.com/Cranial-XIX/Generative-Neural-Parser/blob/693b53dd331e7803600b9be2d23e79719d4603fe/model.py#L605 I noticed you are using LSTM-minus for the inside span of C, which I' not sure is OK. The minus operation in the NN sometime causes issues which results high weights. (because large weights won't be penalized after minus) Is that possible to just use CIv?

wddabc commented 7 years ago

I might consider torch.cat( (AAv, BBv, BIv, CIv, BIv), 1 )