README.md

metadata

dataset_info:
  features:
    - name: Problem
      dtype: string
    - name: Rationale
      dtype: string
    - name: options
      dtype: string
    - name: correct
      dtype: string
    - name: annotated_formula
      dtype: string
    - name: linear_formula
      dtype: string
    - name: category
      dtype: string
  splits:
    - name: train
      num_bytes: 23143391
      num_examples: 29837
    - name: test
      num_bytes: 2313345
      num_examples: 2985
    - name: validation
      num_bytes: 3455493
      num_examples: 4475
  download_size: 12989753
  dataset_size: 28912229
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
      - split: validation
        path: data/validation-*

MMLU dataset translated to Bulgarian

The MMLU dataset translated to Bulgarian language with the opus-mt-tc-big-en-bg model.