---
dataset_info:
  features:
  - name: prompt
    dtype: string
  - name: think
    dtype: string
  - name: response
    dtype: string
  splits:
  - name: train
    num_bytes: 2136931814
    num_examples: 4500000
  download_size: 353234985
  dataset_size: 2136931814
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
license: mit
language:
- en
---