---
dataset_info:
  features:
  - name: instruction
    dtype: string
  - name: input
    struct:
    - name: category
      sequence: string
    - name: interaction
      sequence: string
    - name: user_id
      dtype: string
  splits:
  - name: train
    num_bytes: 26997284
    num_examples: 21306
  download_size: 10700018
  dataset_size: 26997284
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---