---
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
  - split: validation
    path: data/validation-*
license: mit
task_categories:
- zero-shot-classification
language:
- en
tags:
- medical
- biology
- chemistry
- code
dataset_info:
  features:
  - name: dataset
    dtype: string
  - name: id
    dtype: string
  - name: question
    dtype: string
  - name: choices
    sequence: string
  - name: rationale
    dtype: string
  - name: answer
    dtype: string
  splits:
  - name: train
    num_bytes: 92466428
    num_examples: 201676
  - name: test
    num_bytes: 1021843
    num_examples: 2426
  - name: validation
    num_bytes: 1697042
    num_examples: 3345
  download_size: 56075931
  dataset_size: 95185313
---