README.md

metadata

dataset_info:
  - config_name: afrikaans
    features:
      - name: question_text
        dtype: string
      - name: background
        dtype: string
      - name: answer_text
        dtype: string
      - name: options
        sequence: string
      - name: language
        dtype: string
      - name: level
        dtype: string
      - name: subject
        dtype: string
      - name: subject_category
        dtype: string
    splits:
      - name: dev
        num_bytes: 8860
        num_examples: 25
      - name: test
        num_bytes: 194333
        num_examples: 258
    download_size: 71295
    dataset_size: 203193
  - config_name: chinese
    features:
      - name: question_text
        dtype: string
      - name: background
        dtype: string
      - name: answer_text
        dtype: string
      - name: options
        sequence: string
      - name: language
        dtype: string
      - name: level
        dtype: string
      - name: subject
        dtype: string
      - name: subject_category
        dtype: string
    splits:
      - name: dev
        num_bytes: 25055
        num_examples: 29
      - name: test
        num_bytes: 485093
        num_examples: 682
    download_size: 289255
    dataset_size: 510148
  - config_name: english
    features:
      - name: question_text
        dtype: string
      - name: background
        dtype: string
      - name: answer_text
        dtype: string
      - name: options
        sequence: string
      - name: language
        dtype: string
      - name: level
        dtype: string
      - name: subject
        dtype: string
      - name: subject_category
        dtype: string
    splits:
      - name: dev
        num_bytes: 12792
        num_examples: 32
      - name: test
        num_bytes: 2573796
        num_examples: 1911
    download_size: 697219
    dataset_size: 2586588
  - config_name: italian
    features:
      - name: question_text
        dtype: string
      - name: background
        dtype: string
      - name: answer_text
        dtype: string
      - name: options
        sequence: string
      - name: language
        dtype: string
      - name: level
        dtype: string
      - name: subject
        dtype: string
      - name: subject_category
        dtype: string
    splits:
      - name: dev
        num_bytes: 5834
        num_examples: 18
      - name: test
        num_bytes: 2397963
        num_examples: 811
    download_size: 326671
    dataset_size: 2403797
  - config_name: javanese
    features:
      - name: question_text
        dtype: string
      - name: background
        dtype: string
      - name: answer_text
        dtype: string
      - name: options
        sequence: string
      - name: language
        dtype: string
      - name: level
        dtype: string
      - name: subject
        dtype: string
      - name: subject_category
        dtype: string
    splits:
      - name: dev
        num_bytes: 1425
        num_examples: 6
      - name: test
        num_bytes: 187280
        num_examples: 371
    download_size: 84085
    dataset_size: 188705
  - config_name: portuguese
    features:
      - name: question_text
        dtype: string
      - name: background
        dtype: string
      - name: answer_text
        dtype: string
      - name: options
        sequence: string
      - name: language
        dtype: string
      - name: level
        dtype: string
      - name: subject
        dtype: string
      - name: subject_category
        dtype: string
    splits:
      - name: dev
        num_bytes: 20979
        num_examples: 24
      - name: test
        num_bytes: 941655
        num_examples: 889
    download_size: 614816
    dataset_size: 962634
  - config_name: swahili
    features:
      - name: question_text
        dtype: string
      - name: background
        dtype: string
      - name: answer_text
        dtype: string
      - name: options
        sequence: string
      - name: language
        dtype: string
      - name: level
        dtype: string
      - name: subject
        dtype: string
      - name: subject_category
        dtype: string
    splits:
      - name: dev
        num_bytes: 2053
        num_examples: 6
      - name: test
        num_bytes: 607215
        num_examples: 428
    download_size: 94031
    dataset_size: 609268
  - config_name: thai
    features:
      - name: question_text
        dtype: string
      - name: background
        dtype: string
      - name: answer_text
        dtype: string
      - name: options
        sequence: string
      - name: language
        dtype: string
      - name: level
        dtype: string
      - name: subject
        dtype: string
      - name: subject_category
        dtype: string
    splits:
      - name: dev
        num_bytes: 16185
        num_examples: 26
      - name: test
        num_bytes: 2249737
        num_examples: 2168
    download_size: 901256
    dataset_size: 2265922
  - config_name: vietnamese
    features:
      - name: question_text
        dtype: string
      - name: background
        dtype: string
      - name: answer_text
        dtype: string
      - name: options
        sequence: string
      - name: language
        dtype: string
      - name: level
        dtype: string
      - name: subject
        dtype: string
      - name: subject_category
        dtype: string
    splits:
      - name: dev
        num_bytes: 7974
        num_examples: 28
      - name: test
        num_bytes: 767759
        num_examples: 1789
    download_size: 375774
    dataset_size: 775733
configs:
  - config_name: afrikaans
    data_files:
      - split: dev
        path: afrikaans/dev-*
      - split: test
        path: afrikaans/test-*
  - config_name: chinese
    data_files:
      - split: dev
        path: chinese/dev-*
      - split: test
        path: chinese/test-*
  - config_name: english
    data_files:
      - split: dev
        path: english/dev-*
      - split: test
        path: english/test-*
  - config_name: italian
    data_files:
      - split: dev
        path: italian/dev-*
      - split: test
        path: italian/test-*
  - config_name: javanese
    data_files:
      - split: dev
        path: javanese/dev-*
      - split: test
        path: javanese/test-*
  - config_name: portuguese
    data_files:
      - split: dev
        path: portuguese/dev-*
      - split: test
        path: portuguese/test-*
  - config_name: swahili
    data_files:
      - split: dev
        path: swahili/dev-*
      - split: test
        path: swahili/test-*
  - config_name: thai
    data_files:
      - split: dev
        path: thai/dev-*
      - split: test
        path: thai/test-*
  - config_name: vietnamese
    data_files:
      - split: dev
        path: vietnamese/dev-*
      - split: test
        path: vietnamese/test-*

M3Exam: A Multilingual 🌏, Multimodal 🖼, Multilevel 📈 Benchmark for LLMs

This is the repository for M3Exam: A Multilingual, Multimodal, Multilevel Benchmark for Examining Large Language Models.

TL;DR: We introduce M3Exam, a novel benchmark sourced from real and official human exam questions for evaluating LLMs in a multilingual, multimodal, and multilevel context.