Regarding the BIRD benchmark submission, in the instructions it says:
test_databases: The same with dev databases, but with some giant databases.
test_tables.json: The same with dev_tables.json
Does this mean that new databases will appear in the test set, or that the same databases as in the dev set will be used, with identical table structures and metadata, but simply containing more data inside?
Regarding the BIRD benchmark submission, in the instructions it says:
test_databases: The same with dev databases, but with some giant databases.
test_tables.json: The same with dev_tables.json
Does this mean that new databases will appear in the test set, or that the same databases as in the dev set will be used, with identical table structures and metadata, but simply containing more data inside?