Open zhangpengshan opened 9 years ago
Need read paper on ADADELTA, still not clear about the learning rate computation.
Postpone this issue to 0.2.6.
Postpone this issue to 0.2.8. Should learn from stanfordnlp
Check StanfordNLP project in Github.
So far, our learning rate is a fixed value, some paper and mllib are starting to use adaptive learning rate according to current iteration.
This is useful to decrease total iteration number.