Welcome to the competition for the LogicNLG dataset. This task requires model to read a table and generate sentences with logical/numeric inference. More details can be found in LogicNLG website.
In order to match your submissions with the groundtruth, name your submission file as "test_answers.json". The submissions will be evaluated using BLEU-1/2/3 and Adv-Acc socres. The data and evaluation script can be downloaded from here.
This page enumerated the terms and conditions of the competition.
Start: April 16, 2020, midnight
Dec. 31, 2099, midnight
You must be logged in to participate in competitions.Sign In