Why does sample_submission-v0.2.csv have fewer rows than test_public-v0.2.csv?

ESCI Challenge for Improving Product Search
#1

As said in the previous thread, test set is a product recall set. Shouldn’t sample_submission also be a product recall set and have the same number of rows as the test set?

Thank you.

#2

Solved. I just run the provided baseline model and the generated submission file has the same number of rows as the test data. So sample_submission is indeed lacking rows.