Hi,

I am trying to train a syntax-based string to tree German to English
translation model. After tuning, the file "run1.out" contains
sentences in German. Is it normal?

I also tried the hierarchical model and found it the same, but for the
phrase-based model, the file was in English. Decoding the test set
with the new weights, produces English sentences. Also, the BLEU for
the test set is usually higher than for the tuning set and it might be
because of a wrong tuning process. Am I making a mistake in using the
MERT?

Thanks in advance
_______________________________________________
Moses-support mailing list
[email protected]
http://mailman.mit.edu/mailman/listinfo/moses-support

Reply via email to