keyuding / e2emathsolver

Implementation of NAACL 2019 paper "Semantically-Aligned Equation Generation for Solving and Reasoning Math Word Problem"

End-to-End Math Solver

The implementation of the NAACL 2019 paper Semantically-Aligned Equation Generation for Solving and Reasoning Math Word Problems.

Usage

0. Download required files:
   • Math23K dataset. If it is not in the valid JSON format, you need to fix it.
   • Chinese FastText word vectors
1. Preprocess dataset:

cd src/
mkdir ../data/
python make_train_valid.py ~/storage/MWP/data/math23k_fix.json ~/storage/cc.zh.300.vec ../data/train_valid.pkl --valid_ratio 0 --char_based

Note that the warning is not unexpected, because some of the problems use operator out of +, -, *, /. The purpose of the flags are as follow:

• --valid-ratio 0: Set the ratio the validation dataset should take. It should be set to 0 when running 5-fold cross validation.
• --char_based: Specify if you want to tokenize the problem text into characters rather into words.

2. Start 5-fold cross-validation by

mkdir ../models
python train.py ../data/train_valid.pkl ../models/model.pkl --five_fold