[KongchangAI]
Benchmark

UniQL

A text-to-SQL benchmark that translates questions from the BIRD dataset into multiple SQL dialects, used to evaluate model generalization across different database dialects.

Source Articles