group-attention A pytorch implementation of the paper "Modeling Intra-Relation in Math Word Problems with Different Functional Multi-Head Attentions". To run and train this on dataset Math23K, you simply need to : sh scripts/exe_post.sh fold_name_you_wish_to_save