MinGram: A Minimalist Unigram Tokenizer with High Compression and Competitive Morphological Alignment
For NLP practitioners, MinGram offers a simpler, faster Unigram training method that matches or exceeds existing tokenizers in compression and downstream quality.
MinGram simplifies Unigram tokenizer training by replacing complex components with a BPE-derived seed vocabulary, Hard EM, and single pruning step, achieving better compression than BPE and standard Unigram across six languages while maintaining competitive morphological alignment and downstream LM performance.
The Unigram tokenizer uses an elegant representation which makes it straightforward to edit vocabularies, but its training is comparatively heavy and complex. We introduce MinGram (Minimalist Unigram), which keeps the token-list representation but simplifies training using a BPE-derived seed vocabulary, Hard EM on a minimum-token path, and a single flat score-pruning step. This removes the suffix array, the forward-backward pass, and the iterative prune loop, leaving a procedure that requires little beyond tokenizer inference itself. By making token count the primary objective and using a Unigram score only as a tiebreak, MinGram keeps the compression of pure token-count methods while retaining much of the morphological alignment and downstream quality of probabilistic ones. Across six languages, MinGram compresses better than both BPE and standard Unigram, and a compression-oriented variant matches the strongest token-count compressors while retaining substantially higher morphological alignment. In controlled downstream language-model training, Unigram-family tokenizers, with MinGram among the best, consistently beat BPE in bits-per-byte.