---
title: README
emoji: 🐦
colorFrom: pink
colorTo: indigo
sdk: static
pinned: false
---

Hi, I am a magpie 🐦!

🕸️ **Project Website**: [https://magpie-align.github.io/](https://magpie-align.github.io/)

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

🤗 **HF Paper Page**: [https://huggingface.co/papers/2406.08464](https://huggingface.co/papers/2406.08464)

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

🤗 **Magpie Demo**: [https://huggingface.co/spaces/davanstrien/magpie](https://huggingface.co/spaces/davanstrien/magpie) (Thanks a lot for the implementation from @davanstrien!)

🐦 **MagpieLM**: [MagpieLM-4B](https://huggingface.co/spaces/yuchenlin/MagpieLM-4B), [MagpieLM-8B](https://huggingface.co/spaces/yuchenlin/MagpieLM-8B)

**Questions?** Please contact [Zhangchen](mailto:zxu9@uw.edu) and/or [Yuchen](mailto:yuchenl@allenai.org) by email or raise an issue in [Github](https://github.com/magpie-align/magpie/issues/new/choose).

## [🧭 Click here for full dataset navigation (SFT and DPO)](https://github.com/magpie-align/magpie/blob/main/navigation.md)

## Raw Datasets
|Model Name | Dataset | Type | Description |
|-------------|:-------|:-------|:-------|
| [Qwen2.5 72B Instruct](https://huggingface.co/Qwen/Qwen2.5-72B-Instruct) | [Magpie-Qwen2.5-Pro-1M](https://huggingface.co/datasets/Magpie-Align/Magpie-Qwen2.5-Pro-1M-v0.1) | SFT | 1M Raw conversations built with Qwen2.5 72B Instruct.
| [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 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct) | [Magpie-Pro-1M](https://huggingface.co/datasets/Magpie-Align/Llama-3-Magpie-Pro-1M-v0.1) | SFT | 1M Raw conversations built with Meta Llama 3 70B.
| [Llama 3 8B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) | [Magpie-Air-3M](https://huggingface.co/datasets/Magpie-Align/Llama-3-Magpie-Air-3M-v0.1) | SFT | 3M Raw conversations built with Meta Llama 3 8B.
| [Qwen2 72B Instruct](https://huggingface.co/Qwen/Qwen2-72B-Instruct) | [Magpie-Qwen2-Pro-1M](https://huggingface.co/datasets/Magpie-Align/Magpie-Qwen2-Pro-1M-v0.1) | SFT | 1M Raw conversations built with Qwen2 72B Instruct.
| [Qwen2 7B Instruct](https://huggingface.co/Qwen/Qwen2-7B-Instruct) | [Magpie-Qwen2-Air-3M](https://huggingface.co/datasets/Magpie-Align/Magpie-Qwen2-Air-3M-v0.1) | SFT | 3M Raw conversations built with Qwen2 7B Instruct.
| [Phi-3 Medium Instruct](https://huggingface.co/microsoft/Phi-3-medium-128k-instruct) | [Magpie-Phi3-Pro-1M](https://huggingface.co/datasets/Magpie-Align/Magpie-Phi3-Pro-1M-v0.1) | SFT | 1M Raw conversations built with Phi-3 Medium Instruct.
| [Gemma-2-27b-it](https://huggingface.co/google/gemma-2-27b-it) | [Magpie-Gemma2-Pro-534K](https://huggingface.co/datasets/Magpie-Align/Magpie-Gemma2-Pro-534K-v0.1) | SFT | 534K conversations built with Gemma-2-27b-it.
| [Llama 3.1 405B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-405B-Instruct) | [Magpie-Ultra-v0.1](https://huggingface.co/datasets/argilla/magpie-ultra-v0.1) | SFT | [Argilla] 50K Raw conversations built with Meta Llama 3.1 405B.

### Recommended Filtered Datasets

Here are some filtered datasets made by the authors, which are utilized in our [Magpie-Align models](https://huggingface.co/collections/Magpie-Align/magpie-models-668c4a8eea81ccc0db130bdf). We also encourage you to [create and apply your own filters to customize datasets](https://github.com/magpie-align/magpie?tab=readme-ov-file#4-design-and-apply-your-filter). 

We've kept these datasets within the 200K-300K range for your convenience. We found this range represents a sweet spot balancing model performance and training time.

The full list of filtered datasets can be found [here](https://github.com/magpie-align/magpie/blob/main/navigation.md).

|Model Name | Dataset | Size | Type | Description |
|-------------|:-------|:-------|:-------|:-------|
| [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) | 300K | SFT | (🌟 Flexible License! 🌟) Select 300K high quality multi-turn conversations from Magpie-Llama-3.1-Pro-MT-500K.
| [Llama 3 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct) | [Magpie-Pro-300K-Filtered](https://huggingface.co/datasets/Magpie-Align/Magpie-Pro-300K-Filtered) | 300K | SFT | Apply a filter and select 300K high quality conversations from Magpie-Pro-1M.
| [Llama 3 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct) | [Magpie-Pro-MT-300K](https://huggingface.co/datasets/Magpie-Align/Magpie-Pro-MT-300K-v0.1) | 300K | SFT | Select 300K difficult questions from Magpie-Pro-1M and extend to multi-turn conversations.
| [Llama 3 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct) | [Magpie-Reasoning-150K](https://huggingface.co/datasets/Magpie-Align/Magpie-Reasoning-150K) | 150K | SFT | Reasoning booster with 150K math + code + reasoning conversations. Recommend mixing with Magpie-Pro-MT-300K.
| [Qwen2 72B Instruct](https://huggingface.co/Qwen/Qwen2-72B-Instruct) | [Magpie-Qwen2-Pro-200K-Chinese](https://huggingface.co/datasets/Magpie-Align/Magpie-Qwen2-Pro-200K-Chinese) | 200K | SFT | Apply a filter and select 200K high quality Chinese conversations from Magpie-Qwen2-Pro-1M.
| [Gemma-2-27b-it](https://huggingface.co/google/gemma-2-27b-it) | [Magpie-Gemma2-Pro-200K-Filtered](https://huggingface.co/datasets/Magpie-Align/Magpie-Gemma2-Pro-200K-Filtered) | 200K | SFT | (🌟 Flexible License! 🌟) Apply a filter and select 200K conversations from Magpie-Gemma2-Pro-534K.
| [Llama 3 8B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) | [Magpie-Air-DPO-100K](https://huggingface.co/datasets/Magpie-Align/Magpie-Air-DPO-100K-v0.1) | 100K | DPO | DPO dataset via Best-of-N sampling and rewards.