---
dataset_info:
  features:
  - name: system
    dtype: string
  - name: user
    dtype: string
  - name: assistant
    dtype: string
  - name: input_ids
    sequence: int32
  - name: attention_mask
    sequence: int8
  - name: labels
    sequence: int64
  splits:
  - name: train
    num_bytes: 18193906.8
    num_examples: 2763
  - name: test
    num_bytes: 2021545.2
    num_examples: 307
  download_size: 7830262
  dataset_size: 20215452.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---