README.md

---
dataset_info:
  features:
  - name: input
    struct:
    - name: asr
      dtype: string
    - name: input_json
      struct:
      - name: document
        dtype: string
    - name: llm
      dtype: string
    - name: ocr
      dtype: string
    - name: output_type
      dtype: string
    - name: possible_values
      sequence: string
    - name: question
      dtype: string
    - name: strategy
      dtype: string
  - name: label
    dtype: string
  splits:
  - name: train
    num_bytes: 184
    num_examples: 1
  - name: validation
    num_bytes: 184
    num_examples: 1
  download_size: 15285
  dataset_size: 368
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---