I am trying to train the nmt to translate from English to Arabic but it gives errors.
First how should I make the "vocab.ara" and "vocab.en" files is there a script that create these files or its jus the vocabulary from the training dataset?
Second could you give me a fast glimpse why are put at first rows in the vocab files.
I am trying to train the nmt to translate from English to Arabic but it gives errors. First how should I make the "vocab.ara" and "vocab.en" files is there a script that create these files or its jus the vocabulary from the training dataset? Second could you give me a fast glimpse why are put at first rows in the vocab files.