Setting up the Data Printer with Improved English to Ukrainian Machine Translation
This enables faster curation of Ukrainian datasets for large language models, addressing a domain-specific need for Ukrainian NLP.
The paper tackled the problem of expanding Ukrainian language model corpora by developing an improved English-to-Ukrainian machine translation system, achieving state-of-the-art performance on the FLORES devtest set with their decoder-only model Dragoman.
To build large language models for Ukrainian we need to expand our corpora with large amounts of new algorithmic tasks expressed in natural language. Examples of task performance expressed in English are abundant, so with a high-quality translation system our community will be enabled to curate datasets faster. To aid this goal, we introduce a recipe to build a translation system using supervised finetuning of a large pretrained language model with a noisy parallel dataset of 3M pairs of Ukrainian and English sentences followed by a second phase of training using 17K examples selected by k-fold perplexity filtering on another dataset of higher quality. Our decoder-only model named Dragoman beats performance of previous state of the art encoder-decoder models on the FLORES devtest set.