Ground truth for evaluation

[tongjiaming]

Hi, I am trying to evaluate the accuracy of the model, but I found that the evaluation script is only comparing the given prediction and the answer, can you provide the origin data for evaluation?

[celbree]

As a benchmark dataset, we don't release the ground truth in some tasks. You can participant our benchmark leaderboard by sending submissions to codexglue@microsoft.com.