This is an improved grammar correction dataset (English) that includes questions as well, so LLMs trained on this dataset will know how to not reply to questions and instead correct the grammar.
Huge thanks to Parasail for giving me the resources to create this dataset!
