Question About Performance

[cocaer]

The paper shows the best en-fr bleu is 33.4. The readme.md shows
'epoch -> 7
valid_fr-en_mt_bleu -> 28.36
valid_en-fr_mt_bleu -> 30.50
test_fr-en_mt_bleu -> 34.02
test_en-fr_mt_bleu -> 36.62'.
Does this result from the max_len parameter which removes the long sentences from parallel test corpus?

[glample]

No, the difference comes from the fact that the monolingual dataset is different. In the paper we use all NewsCrawl, in the Github we just use NewsCrawl 2013 and 2014 I believe, which is more in domain with newstest2014 on which we evaluate.