---
dataset_info:
  features:
  - name: prompt
    dtype: string
  - name: answer
    dtype: string
  - name: metadata
    dtype: string
  - name: task
    dtype: string
  - name: mode
    dtype: string
  - name: level
    dtype: int64
  splits:
  - name: train
    num_bytes: 6548430546
    num_examples: 7590000
  download_size: 1590387576
  dataset_size: 6548430546
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---