---
dataset_info:
  features:
  - name: input
    dtype: string
  - name: target
    dtype: float64
  splits:
  - name: train
    num_bytes: 155330
    num_examples: 627
  - name: test
    num_bytes: 76756
    num_examples: 269
  download_size: 143023
  dataset_size: 232086
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
how it was yielded: filtered gsm-hard dataset by abs(result) < 1e4. 
---