---
dataset_info:
  features:
  - name: problem
    dtype: string
  - name: answer
    dtype: string
  - name: conversations
    list:
    - name: from
      dtype: string
    - name: value
      dtype: string
  - name: images
    list: image
  - name: _qid
    dtype: int32
  splits:
  - name: train
    num_bytes: 3368835060
    num_examples: 210000
  - name: test
    num_bytes: 1443795618
    num_examples: 90000
  download_size: 4787431789
  dataset_size: 4812630678
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---