SubMerge: Merging Equivalent Subword Tokenizations for Subword Regularized Models in Neural Machine Translation

SubMerge: Merging Equivalent Subword Tokenizations for Subword Regularized Models in Neural Machine Translation Haiyue Song author Francois Meyer author Raj Dabre author Hideki Tanaka author Chenhui Chu author Sadao Kurohashi author 2024-06 text Proceedings of the 25th Annual Conference of the European Association for Machine Translation (Volume 1) Carolina Scarton editor Charlotte Prescott editor Chris Bayliss editor Chris Oakley editor Joanna Wright editor Stuart Wrigley editor Xingyi Song editor Edward Gow-Smith editor Rachel Bawden editor Víctor M Sánchez-Cartagena editor Patrick Cadwell editor Ekaterina Lapshinova-Koltunski editor Vera Cabarrão editor Konstantinos Chatzitheodorou editor Mary Nurminen editor Diptesh Kanojia editor Helena Moniz editor European Association for Machine Translation (EAMT) Sheffield, UK conference publication song-etal-2024-submerge https://aclanthology.org/2024.eamt-1.15/ 2024-06 147 163