Flexing in 73 Languages: A Single Small Model for Multilingual Inflection
This addresses the lack of an open-source, general-purpose multilingual inflection system for handling unseen words across diverse languages, though it is incremental as it builds on existing multilingual modeling techniques.
The authors tackled the problem of generating inflected word forms across multiple languages by developing a compact, single-model approach trained on 73 languages, which outperformed monolingual baselines in most languages and simplified deployment by eliminating the need for separate models.
We present a compact, single-model approach to multilingual inflection, the task of generating inflected word forms from base lemmas to express grammatical categories. Our model, trained jointly on data from 73 languages, is lightweight, robust to unseen words, and outperforms monolingual baselines in most languages. This demonstrates the effectiveness of multilingual modeling for inflection and highlights its practical benefits: simplifying deployment by eliminating the need to manage and retrain dozens of separate monolingual models. In addition to the standard SIGMORPHON shared task benchmarks, we evaluate our monolingual and multilingual models on 73 Universal Dependencies (UD) treebanks, extracting lemma-tag-form triples and their frequency counts. To ensure realistic data splits, we introduce a novel frequency-weighted, lemma-disjoint train-dev-test resampling procedure. Our work addresses the lack of an open-source, general-purpose, multilingual morphological inflection system capable of handling unseen words across a wide range of languages, including Czech. All code is publicly released at: https://github.com/tomsouri/multilingual-inflection.