issai/ARC_Kazakh_Russian
收藏Hugging Face2026-02-18 更新2026-04-05 收录
下载链接:
https://hf-mirror.com/datasets/issai/ARC_Kazakh_Russian
下载链接
链接失效反馈官方服务:
资源简介:
---
dataset_info:
- config_name: kazakh
features:
- name: id
dtype: string
- name: question
dtype: string
- name: choices
struct:
- name: text
list: string
- name: label
list: string
- name: answerKey
dtype: string
- name: source
dtype: string
splits:
- name: test
num_bytes: 1695580
num_examples: 3548
download_size: 711750
dataset_size: 1695580
- config_name: russian
features:
- name: id
dtype: string
- name: question
dtype: string
- name: choices
dtype: string
- name: answerKey
dtype: int64
- name: source
dtype: string
splits:
- name: test
num_bytes: 1845728
num_examples: 3548
download_size: 790394
dataset_size: 1845728
configs:
- config_name: kazakh
data_files:
- split: test
path: kazakh/test-*
- config_name: russian
data_files:
- split: test
path: russian/test-*
license: cc-by-sa-4.0
task_categories:
- question-answering
language:
- kk
---
## Dataset Summary
These are the machine-translated Kazakh and Russian versions of the [ARC (AI2 Reasoning Challenge)](https://huggingface.co/datasets/allenai/ai2_arc) dataset (test set).
ARC is specifically designed to focus on genuine reasoning abilities, containing elementary and middle-school science questions that require more than simple retrieval or word-matching to solve. These Kazakh and Russian versions serve as a benchmark for evaluating how well models can perform complex reasoning and apply scientific knowledge when prompted in Kazakh or Russian.
数据集信息:
- 配置名称:哈萨克语(Kazakh)
特征:
- 字段名:id,数据类型:字符串(string)
- 字段名:question,数据类型:字符串(string)
- 字段名:choices,结构体(struct)类型,包含两个成员:
- 成员名:text,类型:字符串列表(list[string])
- 成员名:label,类型:字符串列表(list[string])
- 字段名:answerKey,数据类型:字符串(string)
- 字段名:source,数据类型:字符串(string)
数据集划分:
- 划分名称:测试集(test),字节数:1695580,样本数:3548
下载大小:711750,数据集总大小:1695580
- 配置名称:俄语(Russian)
特征:
- 字段名:id,数据类型:字符串(string)
- 字段名:question,数据类型:字符串(string)
- 字段名:choices,数据类型:字符串(string)
- 字段名:answerKey,数据类型:64位整数(int64)
- 字段名:source,数据类型:字符串(string)
数据集划分:
- 划分名称:测试集(test),字节数:1845728,样本数:3548
下载大小:790394,数据集总大小:1845728
配置项:
- 配置名称:哈萨克语(Kazakh),数据文件:
- 划分:测试集(test),路径:kazakh/test-*
- 配置名称:俄语(Russian),数据文件:
- 划分:测试集(test),路径:russian/test-*
许可证:CC BY-SA 4.0(知识共享署名-相同方式共享4.0协议)
任务类别:问答任务(Question Answering)
语言:哈萨克语(kk)、俄语(ru)
## 数据集概览
本数据集为[ARC(AI2推理挑战,AI2 Reasoning Challenge)](https://huggingface.co/datasets/allenai/ai2_arc)数据集(测试集)的机器翻译哈萨克语与俄语版本。
ARC数据集专门聚焦真实推理能力的测评,涵盖中小学科学类考题,解决此类问题需超越简单的信息检索或词汇匹配范畴。此次发布的哈萨克语与俄语版本可作为基准测试集,用于评估模型在以哈萨克语或俄语作为提示语言时,完成复杂推理并应用科学知识的能力。
提供机构:
issai


