---
dataset_info:
  features:
  - name: prompt
    dtype: string
  - name: operand_len_a
    dtype: int64
  - name: operand_len_b
    dtype: int64
  - name: answer
    dtype: string
  splits:
  - name: train
    num_bytes: 273876912
    num_examples: 4500000
  - name: val
    num_bytes: 30430768
    num_examples: 500000
  - name: val_extrapolate
    num_bytes: 19417250
    num_examples: 100000
  - name: test
    num_bytes: 19417236
    num_examples: 100000
  download_size: 211112489
  dataset_size: 343142166
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: val
    path: data/val-*
  - split: val_extrapolate
    path: data/val_extrapolate-*
  - split: test
    path: data/test-*
---