---
dataset_info:
  features:
  - name: messages
    list:
    - name: role
      dtype: string
    - name: content
      dtype: string
  - name: slice
    dtype: string
  - name: source
    dtype: string
  - name: category
    dtype: string
  - name: has_tools
    dtype: bool
  - name: n_turns
    dtype: int64
  splits:
  - name: train
    num_bytes: 481226694
    num_examples: 40401
  - name: validation
    num_bytes: 53999517
    num_examples: 4489
  download_size: 531789760
  dataset_size: 535226211
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---