You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Comparing with training of using corpus of 2M ch-en, when I use corpus of 940k ch-en to train model, what parameters should I use ?I have tried to use batch_size=25k, learning rate=1.0 and batch_size=25k, learning rate=0.5,but just got bleu=3.6 and bleu=33.7 in MT06,respectively
The text was updated successfully, but these errors were encountered:
Do you initialize the document-level model with the sentence-level model? The default parameter setting in the code is OK. Just make sure you follow the steps in README
Comparing with training of using corpus of 2M ch-en, when I use corpus of 940k ch-en to train model, what parameters should I use ?I have tried to use batch_size=25k, learning rate=1.0 and batch_size=25k, learning rate=0.5,but just got bleu=3.6 and bleu=33.7 in MT06,respectively
The text was updated successfully, but these errors were encountered: