---
license: cc-by-4.0
task_categories:
- text-classification
language:
- an
dataset_info:
  features:
  - name: id
    dtype: string
  - name: premise
    dtype: string
  - name: choice1
    dtype: string
  - name: choice2
    dtype: string
  - name: question
    dtype: string
  - name: label
    dtype: int64
  splits:
  - name: validation
    num_examples: 100
  - name: test
    num_examples: 500
  download_size: 55691
  dataset_size: 77039
configs:
- config_name: default
  data_files:
  - split: validation
    path: validation.tsv
  - split: test
    path: test.tsv
size_categories:
- n<1K
---

---

# **COPA-AN**: Aragonese version of the XCOPA dataset

<img src="https://huggingface.co/datasets/cadiera/copa-an/resolve/main/cadiera_banner.png" width="480">

## *A dataset from Cadiera: benchmarks for the Aragonese language*

---

Aragonese version of the validation and test sets of the XCOPA dataset.

600 instances (500 validation + 100 test), with the following format:

| id | premise | choice1 | choice2 | question | label |
| --- | --- | --- | --- | --- | --- |
| 1 | La nina trobó un cuco en os suyos cereals. | Abocó leit en o cuezo. | Perdió la gana. | effect | 1 |