List your company

Sentence Retrieval on 6 languages Same script split

43.9AccuracyUnigram
27.36431.65735.9540.243May 26, 2023

Evaluation results

MethodLinks
UnigramTokenizer=Unigram2023.0543.9
TokMixTokenizer=TokMix2023.0543.4
BPETokenizer=BPE2023.0540.7
NoOverlapTokenizer=NoOverlap2023.0528