Geographical Distance Is The New Hyperparameter: A Case Study Of Finding The Optimal Pre-trained Language For English-isiZulu Machine Translation

1 code implementation • 17 May 2022 • Muhammad Umair Nasir, Innocent Amos Mchechesi

Stemming from the limited availability of datasets and textual resources for low-resource languages such as isiZulu, there is a significant need to be able to harness knowledge from pre-trained models to improve low resource machine translation.

Low-Resource Neural Machine Translation Transfer Learning +1