Models and datasets from the paper: "LLMs to Replace Crowdsourcing For Parallel Data Creation? The Case of Text Detoxification" by Moskovskiy et al.