1 paper
Go Inoue, Bashar Alhafni, Nurpeiis Baimukan +2
In this paper, we explore the effects of language variants, data sizes, and fine-tuning task types in Arabic pre-trained language models. To do so, we build three pre-trained langu…