mizo-mt-marked

Mizo to English, entity-marked source. BLEU 49.53

Trained on the Mizo NER corpus, 441,178 sentences with silver-standard labels.

Usage

from transformers import pipeline
mt = pipeline("translation", model="haulai/mizo-mt-marked")
print(mt("Kar thum chawlh Liana-an a la."))

Important

The entity-marked model expects markup produced by the projection pipeline. Feeding it markup from a recognizer removes the BLEU gain and lowers entity retention — see the paper.

Trained only on entity-bearing sentences, so behaviour on entity-sparse text is untested.

Citation

See the code repository.