UKPLab/dara
收藏Hugging Face2024-08-01 更新2025-04-12 收录
下载链接：
https://hf-mirror.com/datasets/UKPLab/dara
下载链接
链接失效反馈官方服务：
资源简介：
---
dataset_info:
features:
- name: input
dtype: string
- name: output
sequence: string
- name: program
dtype: string
splits:
- name: train
num_bytes: 2898123
num_examples: 768
download_size: 820960
dataset_size: 2898123
configs:
- config_name: default
data_files:
- split: train
path: data/train-00000-of-00001.parquet
---
# Dataset Card for DARA
## Dataset Summary
This dataset contains 768 curated reasoning trajectories used to train DARA (https://github.com/UKPLab/acl2024-DARA). It is sourced from GrailQA, WebQSP, and GraphQ.
The fields include:
- Input: Question
- Output: The stepwise reasoning trajectory to construct the full logical form step by step.
- Program: Broken-down subtasks based on the original logical form.
## Uses
```python
from datasets import load_dataset
dataset = load_dataset("UKPLab/dara")
```
## Additional Information
- **Repository**: https://github.com/UKPLab/acl2024-DARA
- **Paper**: https://arxiv.org/abs/2406.07080
提供机构：
UKPLab
