QARV / README.md
amphora's picture
Update README.md
babec4b verified
---
dataset_info:
features:
- name: Question
dtype: string
- name: Source
dtype: string
splits:
- name: train
num_bytes: 7971
num_examples: 104
download_size: 5407
dataset_size: 7971
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---
# QARV (Question and Answers with Regional Variance)
```EleutherAI Community Project```
The QARV (Question and Answers with Regional Variance) project aims to curate a collection of questions with answers that exhibit regional variations across different nations.
## Version
This is the first slice of the QARV dataset collected by (1) manual collection by humans or (2) GPT-4 Generation with RAG (wikipedia) and human filtering.
Collected questions undergo a second quality filtering (removal of inadequate questions and Yes/No questions) and deduplicaton.