PKU-Alignment/Align-Anything-Instruction-100K

Align-Anything-Instruction-100K数据集是一个高质量的指令遵循数据集，包含105,333个问答对。这些问答对是通过GPT-4精心注释和提炼的指令生成的。数据集的提示来源于多个公开数据集，如PKU-SafeRLHF QA、DialogSum、Empathetic、Alpaca和InstructionWild。每个提示都在专家示范和特定指导下由GPT-4进行提炼，并由GPT-4对响应进行注释，从而形成了一个高质量的指令遵循数据集。数据集在多种任务类型上具有广泛的覆盖，并在Just-Eval基准上展示了优秀的性能。

Align-Anything-Instruction-100K is a high-quality instruction-following dataset consisting of 100K question-answer entries, annotated and refined by GPT-4. The prompts are sourced from multiple public datasets such as PKU-SafeRLHF Dataset QA, DialogSum, Empathetic Dataset, Alpaca, and InstructionWild. Each prompt is refined by GPT-4 under expert demonstration and specific guidelines, followed by GPT-4s annotation of the responses. The dataset covers a broad range of prompt types and includes various task types such as text summarization, sentiment analysis, etc. Additionally, the dataset has been evaluated on the Just-Eval benchmark, assessing the responses across five dimensions: helpfulness, clarity, factuality, depth, and engagement.