---
configs:
- config_name: ar
  data_files:
  - split: test
    path: ar/test-*
- config_name: default
  data_files:
  - split: test
    path: data/test-*
- config_name: en
  data_files:
  - split: test
    path: en/test-*
- config_name: fr
  data_files:
  - split: test
    path: fr/test-*
- config_name: ru
  data_files:
  - split: test
    path: ru/test-*
dataset_info:
- config_name: ar
  features:
  - name: sentence1
    dtype: string
  - name: sentence2
    dtype: string
  - name: score
    dtype: float64
  splits:
  - name: test
    num_bytes: 158273
    num_examples: 30
  download_size: 88385
  dataset_size: 158273
- config_name: default
  features:
  - name: sentence1
    dtype: string
  - name: sentence2
    dtype: string
  - name: score
    dtype: float64
  - name: lang
    dtype: string
  splits:
  - name: test
    num_bytes: 710237
    num_examples: 120
  download_size: 434609
  dataset_size: 710237
- config_name: en
  features:
  - name: sentence1
    dtype: string
  - name: sentence2
    dtype: string
  - name: score
    dtype: float64
  splits:
  - name: test
    num_bytes: 178597
    num_examples: 30
  download_size: 119948
  dataset_size: 178597
- config_name: fr
  features:
  - name: sentence1
    dtype: string
  - name: sentence2
    dtype: string
  - name: score
    dtype: float64
  splits:
  - name: test
    num_bytes: 190875
    num_examples: 30
  download_size: 128978
  dataset_size: 190875
- config_name: ru
  features:
  - name: sentence1
    dtype: string
  - name: sentence2
    dtype: string
  - name: score
    dtype: float64
  splits:
  - name: test
    num_bytes: 181772
    num_examples: 30
  download_size: 104390
  dataset_size: 181772
---
# STS22 Human Subset

Unified multilingual dataset with lang column.

Configs:
- default: All languages combined (with lang column)
- en/ar/fr/ru: Individual language configs (no lang column)

Each config has a 'test' split.