README.md

---
dataset_info:
  features:
  - name: id
    dtype: string
  - name: images
    list:
    - name: bytes
      dtype: 'null'
    - name: path
      dtype: string
  - name: conversation
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  splits:
  - name: train
    num_bytes: 32321172
    num_examples: 23201
  - name: eval
    num_bytes: 8707158
    num_examples: 6249
  - name: test
    num_bytes: 19116592
    num_examples: 15480
  download_size: 2724018
  dataset_size: 60144922
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: eval
    path: data/eval-*
  - split: test
    path: data/test-*
---