Andargachew Mekonnen Gezmu
Andargachew Mekonnen Gezmu
Yes, it is to break down the root (stem) and affix structure.
Can the output of the morphophonemic model (the word stem and the meta-morphemes of an orthographic word) be considered as the "morphological segmentation" of the word?
Yes, I want to obtain surface form segmentation. For example, for the word "arabaları" -- "araba+lar+ı".
In your code base, I couldn't get the dedicated morpheme segmentation grammar that marks morpheme boundaries (e.g. evlerinde -- ev+ler+i+nde or arabaları -- araba+lar+ı). Could you please give me any...
Thanks @ozturel for your reply. I hope you will add this functionality soon. Besides, I appreciate that you have restored the stems of some words which have changed through phonological...
SacreBLEU (the current version, v2.0.0) has such implementations for BLUE, chrF, and Translation error rate (TER) metrics. I am not sure whether they are implemented purely in python.
When calculating the BLEU-score with the built-in t2t-bleu tool, the test data sets should be detokenized. In that case, the score will even be lower than the reported 28.54 (cased)....