StaQC (Stack Overflow Question-Code pairs) is the largest dataset to date of around 148K Python and 120K SQL domain question-code pairs, which are automatically mined from Stack Overflow using a Bi-View Hierarchical Neural Network, as described in the paper "StaQC: A Systematically Mined Question-Code Dataset from Stack Overflow" (WWW'18).
For more details, please check https://github.com/LittleYUYU/StackOverflow-Question-Code-Dataset