A problem about data

[jiusigua]

@potsawee Hi, I want to run the code of this project. Can you provide the data set that you use？Or where should I download them？
include source.txt target.txt glove.6B.200d.txt
Thanks very much！

[potsawee]

Hi there, sorry for responding a few months late.

(1) You could download data sets that are publicly available such as FCE, e.g. from https://www.cl.cam.ac.uk/research/nl/bea2019st/#data. You may need to convert an m2 file to source (erroneous) and target (error-corrected) files for training data.

(2) For pre-trained embedding, e.g. GloVe, you can get one from https://nlp.stanford.edu/projects/glove/ or check out Google pre-trained embedding. Note that this RNN-based sequence-to-sequence model is quite obsolete (as in 2019 but it should be good enough as a baseline system), and you may want to have a look at transformer-based systems for more advanced ones.