Dolci-RLZero-General-7B

# Dolci-RL-Zero-General-7B

## Dataset Summary

**Dolci-RL-Zero-General-7B** is the reinforcement learning dataset used to train the *Olmo3-RL-Zero-7B-General* model.
It contains **12,841** general chat prompts sampled from the larger Dolci-Think-RL mixture. The reward was dervied by using an LM judge.

---
## Downloading

You can download and load this data using HuggingFace's `datasets` library with the following code:

```python
from datasets import load_dataset
dataset = load_dataset("allenai/Dolci-RL-Zero-General-7B", split="train",)
```

### Licensing Information

This dataset is licensed under ODC-BY. It is intended for research and educational use in accordance with Ai2's [Responsible Use Guidelines](https://allenai.org/responsible-use).

### Citation
Technical manuscript coming soon!

# Dolci-RL-Zero-General-7B

## 数据集概述

**Dolci-RL-Zero-General-7B** 是用于训练*Olmo3-RL-Zero-7B-General*模型的强化学习（Reinforcement Learning）数据集。
该数据集包含**12,841**条通用对话提示词，均从规模更大的Dolci-Think-RL混合数据集中采样得到。其奖励信号由语言模型评判器（LM Judge）生成。

---

## 下载方式

可通过HuggingFace的`datasets`库下载并加载该数据集，代码如下：

python
from datasets import load_dataset
dataset = load_dataset("allenai/Dolci-RL-Zero-General-7B", split="train",)

### 许可信息

本数据集采用ODC-BY许可证发布，仅供研究与教育用途，需遵循AI2的《负责任使用指南》（https://allenai.org/responsible-use）。

### 引用信息

技术稿件即将发布！