Dataset contains 90k conversation collected in ShareGPT website. Used to finetune LLM to act like chatGPT
