This is a dataset of practical LLM instructions across tasks that tasks we actually use these models for such as coding, reasoning, math, writing, etc.