skandermoalla/qrpo-paper-mistral-sft-ultrafeedback-armorm-temp1-ref50-offpolicy2random-armorm

---
license: mit
tags:
- reinforcement-learning
- alignment
- qrpo
---

# qrpo-paper-mistral-sft-ultrafeedback-armorm-temp1-ref50-offpolicy2random-armorm

Dataset with reference completions and rewards for a specific model and reward model, ready for training with the **QRPO reference codebase** (https://github.com/CLAIRE-Labo/quantile-reward-policy-optimization).

Part of the dataset collection for the paper **Quantile Reward Policy Optimization: Alignment with Pointwise Regression and Exact Partition Functions** (https://arxiv.org/pdf/2507.08068).

---
许可证：MIT
标签：
- 强化学习
- 对齐
- QRPO
---

# qrpo-paper-mistral-sft-ultrafeedback-armorm-temp1-ref50-offpolicy2random-armorm

该数据集包含针对特定模型与奖励模型的参考补全结果及奖励值，可直接用于**QRPO（Quantile Reward Policy Optimization）参考代码库**（https://github.com/CLAIRE-Labo/quantile-reward-policy-optimization）的训练。

本数据集是论文**《分位数奖励策略优化：基于点态回归与精确配分函数的对齐方法》**（Quantile Reward Policy Optimization: Alignment with Pointwise Regression and Exact Partition Functions，https://arxiv.org/pdf/2507.08068）数据集集合的一部分。