---
dataset_info:
  features:
  - name: image
    list:
    - name: bytes
      dtype: binary
    - name: path
      dtype: 'null'
  - name: text
    list:
    - name: assistant
      dtype: string
    - name: source
      dtype: string
    - name: user
      dtype: string
  splits:
  - name: train
    num_bytes: 789039550
    num_examples: 2000
  download_size: 788860093
  dataset_size: 789039550
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---