README.md

metadata

dataset_info:
  - config_name: medium
    features:
      - name: subject
        dtype: string
      - name: question
        dtype: string
      - name: question_en
        dtype: string
      - name: choices
        sequence: string
      - name: choices_en
        sequence: string
      - name: answer
        dtype: int64
    splits:
      - name: test
        num_bytes: 4162480
        num_examples: 7040
    download_size: 1703853
    dataset_size: 4162480
  - config_name: small
    features:
      - name: subject
        dtype: string
      - name: question
        dtype: string
      - name: question_en
        dtype: string
      - name: choices
        sequence: string
      - name: choices_en
        sequence: string
      - name: answer
        dtype: int64
    splits:
      - name: test
        num_bytes: 1554014
        num_examples: 3302
    download_size: 561076
    dataset_size: 1554014
configs:
  - config_name: medium
    data_files:
      - split: test
        path: medium/test-*
  - config_name: small
    data_files:
      - split: test
        path: small/test-*
  - config_name: large
    data_files:
      - split: test
        path:
          - medium/test-*
          - small/test-*

Tamil ARIVU (AI Reasoning and Intelligent Verbal Understanding) Benchmark

Tamil ARIVU is a comprehensive benchmark dataset designed to evaluate the performance of AI models in understanding and reasoning with the Tamil language. The dataset consists of a diverse set of multiple-choice questions (MCQs) sourced from various domains, including but not limited to general knowledge, history, literature, grammar, and culture.