|
|
--- |
|
|
license: mit |
|
|
datasets: |
|
|
- liuwenhan/reasonrank_data_sft |
|
|
- liuwenhan/reasonrank_data_rl |
|
|
- liuwenhan/reasonrank_data_13k |
|
|
language: |
|
|
- en |
|
|
base_model: |
|
|
- Qwen/Qwen2.5-32B-Instruct |
|
|
library_name: transformers |
|
|
tags: |
|
|
- passage-ranking |
|
|
- text-ranking |
|
|
- reasoning |
|
|
- Information-retrieval |
|
|
--- |
|
|
|
|
|
**📢 Update: On September 4, 2025, we merged the LoRA parameters of ReasonRank (32B) into the model’s checkpoint, so now everyone only needs to load the model without having to load the LoRA adapter anymore.** |
|
|
|
|
|
## Introduction |
|
|
|
|
|
This is the model trained in our paper: ReasonRank: Empowering Passage Ranking with Strong Reasoning Ability ([📝arXiv](https://arxiv.org/abs/2508.07050)). Please refer our [🧩github repository](https://github.com/8421BCD/ReasonRank) for the usage of reasonrank-32B. |
|
|
|
|
|
|
|
|
## Model Performance |
|
|
<p align="center"> |
|
|
<img width="90%" alt="image" src="https://8421bcd.oss-cn-beijing.aliyuncs.com/img/image-20250810163757771.png" /> |
|
|
</p> |
|
|
|
|
|
🌹 If you use this model, please ✨star our <a href="https://github.com/8421BCD/reasonrank" target="_blank">GitHub repository</a> to support us. Your star means a lot! |
