Using pretrained En-De model?


(Zewei Chu) #1

I am using the pretrained En-De model with the following command,

python -model …/opennmt-pretrained/ -src data/train-context.txt -output data-translated/train-context-de.txt -replace_unk -verbose

However, the generated results look very weird. I am wondering if I am using the wrong command or missed some parameter to generate the translations. Any advice?

(Guillaume Klein) #2

Did you preprocess the data using the SentencePiece model included in the model archive? This has been discussed many times on the forum.