hi, sorry it's me again :) . During dual training, there isn't change output model. if you have time, see my output log. (log.txt) First i left off evaluation scripts and train dual.Second using evaluation scripts alone, evalute last output dualmodel. But output of translate.py is like this
hit.en.tok.shuf.dev.tok.output.en_fr_1.dev.txt Then BLEU=0.00
any advance?tnx.
hi, sorry it's me again :) . During dual training, there isn't change output model. if you have time, see my output log. (log.txt) First i left off evaluation scripts and train dual.Second using evaluation scripts alone, evalute last output dualmodel. But output of translate.py is like this hit.en.tok.shuf.dev.tok.output.en_fr_1.dev.txt Then BLEU=0.00 any advance?tnx.