Magpie-Llama-3.1-Pro-MT-500K-v0.1


Project Web: [https://magpie-align.github.io/](https://magpie-align.github.io/)

Arxiv Technical Report: [https://arxiv.org/abs/2406.08464](https://arxiv.org/abs/2406.08464)

Codes: [https://github.com/magpie-align/magpie](https://github.com/magpie-align/magpie)

## Abstract
<details><summary>Click Here</summary>
High-quality instruction data is critical for aligning large language models (LLMs). Although some models, such as Llama-3-Instruct, have open weights, their alignment data remain private, which hinders the democratization of AI. High human labor costs and a limited, predefined scope for prompting prevent existing open-source data creation methods from scaling effectively, potentially limiting the diversity and quality of public alignment datasets. Is it possible to synthesize high-quality instruction data at scale by extracting it directly from an aligned LLM? We present a self-synthesis method for generating large-scale alignment data named Magpie. Our key observation is that aligned LLMs like Llama-3-Instruct can generate a user query when we input only the left-side templates up to the position reserved for user messages, thanks to their auto-regressive nature. We use this method to prompt Llama-3-Instruct and generate 4 million instructions along with their corresponding responses. We perform a comprehensive analysis of the extracted data and select 300K high-quality instances. To compare Magpie data with other public instruction datasets, we fine-tune Llama-3-8B-Base with each dataset and evaluate the performance of the fine-tuned models. Our results indicate that in some tasks, models fine-tuned with Magpie perform comparably to the official Llama-3-8B-Instruct, despite the latter being enhanced with 10 million data points through supervised fine-tuning (SFT) and subsequent feedback learning. We also show that using Magpie solely for SFT can surpass the performance of previous public datasets utilized for both SFT and preference optimization, such as direct preference optimization with UltraFeedback. This advantage is evident on alignment benchmarks such as AlpacaEval, ArenaHard, and WildBench.
</details><be>

## Dataset Details
This dataset is generated by [Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct) using [Magpie](https://huggingface.co/Magpie-Align). Please refer to our [paper](https://arxiv.org/abs/2406.08464) and [codebase](https://github.com/magpie-align/magpie) for implementation details.

**License**: Please follow [Meta Llama 3.1 Community License](https://github.com/meta-llama/llama-models/blob/main/models/llama3_1/LICENSE).

### Available Labels
- **Input Length**: The total number of characters in the instructions.
- **Output Length**: The total number of characters in the responses.
- **Task Category**: The specific category of the instructions.
- **Input Quality**: The clarity, specificity, and coherence of the instructions, rated as 'very poor', 'poor', 'average', 'good', and 'excellent'.
- **Input Difficulty**: The level of knowledge required to address the task described in the instruction, rated as 'very easy', 'easy', 'medium', 'hard', or 'very hard'.
- **Minimum Neighbor Distance**: The embedding distance to the nearest neighbor within the dataset. It can be used for filtering out repetitive or similar instances.
- **Safety**: Safety tags marked by [meta-llama/Meta-Llama-Guard-2-8B](https://huggingface.co/meta-llama/Meta-Llama-Guard-2-8B)
- **Reward**: The output of the reward model given the specific instruction-response pair.
- **Language**: The language of the instruction.

## Filter Setups

To create this multi-turn dataset, we filtered [Magpie-Llama-3.1-Pro-1M](https://huggingface.co/datasets/Magpie-Align/Magpie-Llama-3.1-Pro-1M-v0.1) with the following setups:
- **Input Quality**: >= good
- **Instruction Reward**: >=-10
- Remove repetition and incomplete instructions (e.g., end with :)
- Choose instructions with `\n`<5 except for coding & debugging
- Choose 500K data with the longest responses

We then extend them to multi-turn conversations.

## Dataset Navigation 🧭
|Model Name | Dataset | Type | Description |
|-------------|:-------|:-------|:-------|
| [Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct) | [Magpie-Llama-3.1-Pro-1M](https://huggingface.co/datasets/Magpie-Align/Magpie-Llama-3.1-Pro-1M-v0.1) | SFT | 1M Raw conversations built with Meta Llama 3.1 70B.
| [Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct) | [Magpie-Llama-3.1-Pro-300K-Filtered](https://huggingface.co/datasets/Magpie-Align/Magpie-Llama-3.1-Pro-300K-Filtered) | SFT | Apply a filter and select 300K high quality conversations.
| [Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct) | [Magpie-Llama-3.1-Pro-500K-Filtered](https://huggingface.co/datasets/Magpie-Align/Magpie-Llama-3.1-Pro-500K-Filtered) | SFT | Apply a filter and select 500K high quality conversations.
| [Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct) | [Magpie-Llama-3.1-Pro-MT-500K](https://huggingface.co/datasets/Magpie-Align/Magpie-Llama-3.1-Pro-MT-500K-v0.1) | SFT | Extend Magpie-Llama-3.1-Pro-500K-Filtered to multi-turn.
| [Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct) | [Magpie-Llama-3.1-Pro-MT-300K-Filtered](https://huggingface.co/datasets/Magpie-Align/Magpie-Llama-3.1-Pro-MT-300K-Filtered) | SFT | Select 300K high quality multi-turn conversations from Magpie-Llama-3.1-Pro-MT-500K.
| [Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct) | [Magpie-Llama-3.1-Pro-DPO-100K](https://huggingface.co/datasets/Magpie-Align/Magpie-Llama-3.1-Pro-DPO-100K-v0.1) | DPO | DPO dataset via Best-of-N sampling and rewards.


项目网页：[https://magpie-align.github.io/](https://magpie-align.github.io/)

arXiv技术报告：[https://arxiv.org/abs/2406.08464](https://arxiv.org/abs/2406.08464)

代码仓库：[https://github.com/magpie-align/magpie](https://github.com/magpie-align/magpie)

## 摘要
<details><summary>点击此处展开</summary>
高质量指令数据对大语言模型(LLM)的对齐至关重要。尽管部分模型（如Llama-3-Instruct）开放了模型权重，但其对齐数据仍处于私有状态，这阻碍了人工智能的民主化进程。现有开源数据构建方法面临人力成本高昂、提示范围受限且预定义等问题，难以实现有效规模化扩展，可能限制了公开对齐数据集的多样性与质量。能否直接从已对齐的大语言模型中提取信息，从而规模化合成高质量指令数据？我们提出了一种用于生成大规模对齐数据的自合成方法，命名为喜鹊（Magpie）。我们的核心观察是：得益于自回归(auto-regressive)特性，当我们仅输入至用户消息预留位置的左侧模板时，像Llama-3-Instruct这类已对齐的大语言模型能够生成用户查询。我们利用该方法对Llama-3-Instruct进行提示，生成了400万条指令及其对应的响应。我们对提取得到的数据进行了全面分析，并筛选出30万个高质量样本。为了将Magpie数据集与其他公开指令数据集进行对比，我们分别使用每个数据集对Llama-3-8B-Base进行微调，并评估微调后模型的性能。实验结果表明，在部分任务中，使用Magpie数据集微调得到的模型性能可与官方Llama-3-8B-Instruct相媲美——尽管后者通过监督微调(SFT)与后续反馈学习使用了1000万条数据进行增强。我们还证实，仅使用Magpie数据集进行监督微调，其性能便可超越此前同时用于监督微调与偏好优化的公开数据集（如结合UltraFeedback的直接偏好优化(DPO)）。这一优势在AlpacaEval、ArenaHard与WildBench等对齐基准测试中体现明显。
</details><be>

## 数据集详情
本数据集由[Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct)通过[Magpie](https://huggingface.co/Magpie-Align)生成。有关实现细节，请参阅我们的[论文](https://arxiv.org/abs/2406.08464)与[代码库](https://github.com/magpie-align/magpie)。

**许可证**：请遵循[Meta Llama 3.1社区许可证](https://github.com/meta-llama/llama-models/blob/main/models/llama3_1/LICENSE)。

### 可用标签
- **输入长度**：指令的总字符数。
- **输出长度**：响应的总字符数。
- **任务类别**：指令所属的具体类别。
- **输入质量**：指令的清晰度、特异性与连贯性，评级分为"极差"、"较差"、"中等"、"良好"与"优秀"。
- **输入难度**：完成指令描述任务所需的知识水平，评级分为"极简单"、"简单"、"中等"、"困难"与"极困难"。
- **最小邻域距离**：数据集中与当前样本最近邻的嵌入距离，可用于过滤重复或相似样本。
- **安全性**：由[meta-llama/Meta-Llama-Guard-2-8B](https://huggingface.co/meta-llama/Meta-Llama-Guard-2-8B)标记的安全标签。
- **奖励得分**：针对特定指令-响应对的奖励模型输出值。
- **语言**：指令所使用的语言。

## 过滤设置
为构建该多轮对话数据集，我们基于[Magpie-Llama-3.1-Pro-1M](https://huggingface.co/datasets/Magpie-Align/Magpie-Llama-3.1-Pro-1M-v0.1)进行了如下过滤：
- **输入质量**：≥良好
- **指令奖励得分**：≥-10
- 移除重复与不完整的指令（例如以冒号结尾的样本）
- 选取换行符`
`数量小于5的指令（编码与调试任务除外）
- 选取响应长度最长的50万条数据

随后我们将这些样本扩展为多轮对话。

## 数据集导航 🧭
| 模型名称 | 数据集 | 类型 | 描述 |
|-------------|:-------|:-------|:-------|
| [Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct) | [Magpie-Llama-3.1-Pro-1M](https://huggingface.co/datasets/Magpie-Align/Magpie-Llama-3.1-Pro-1M-v0.1) | 监督微调(SFT) | 基于Meta Llama 3.1 70B构建的100万条原始对话数据。
| [Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct) | [Magpie-Llama-3.1-Pro-300K-Filtered](https://huggingface.co/datasets/Magpie-Align/Magpie-Llama-3.1-Pro-300K-Filtered) | 监督微调(SFT) | 经过过滤后选取的30万个高质量对话数据。
| [Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct) | [Magpie-Llama-3.1-Pro-500K-Filtered](https://huggingface.co/datasets/Magpie-Align/Magpie-Llama-3.1-Pro-500K-Filtered) | 监督微调(SFT) | 经过过滤后选取的50万个高质量对话数据。
| [Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct) | [Magpie-Llama-3.1-Pro-MT-500K](https://huggingface.co/datasets/Magpie-Align/Magpie-Llama-3.1-Pro-MT-500K-v0.1) | 监督微调(SFT) | 将Magpie-Llama-3.1-Pro-500K-Filtered扩展为多轮对话形式。
| [Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct) | [Magpie-Llama-3.1-Pro-MT-300K-Filtered](https://huggingface.co/datasets/Magpie-Align/Magpie-Llama-3.1-Pro-MT-300K-Filtered) | 监督微调(SFT) | 从Magpie-Llama-3.1-Pro-MT-500K中选取的30万个高质量多轮对话数据。
| [Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct) | [Magpie-Llama-3.1-Pro-DPO-100K](https://huggingface.co/datasets/Magpie-Align/Magpie-Llama-3.1-Pro-DPO-100K-v0.1) | 直接偏好优化(DPO) | 通过最优样本采样(Best-of-N)与奖励机制构建的DPO数据集。