We have updated the starter-kits and evaluator to pass the path of the
product_catalogue in your
predict function implementation. For example, for task-1, the
predict function is changed to the following : https://gitlab.aicrowd.com/aicrowd/challenges/amazon-kdd-cup/task_1_query-product_ranking_code_starter_kit/-/blob/main/run.py#L20.
Additionally, we are happy to increase the per-test-set timeout for all the three tasks to
Best of Luck,