README.md

---
dataset_info:
  features:
  - name: messages
    list:
    - name: role
      dtype: string
    - name: content
      list:
      - name: text
        dtype: string
      - name: type
        dtype: string
  - name: images
    sequence: 'null'
  - name: problem
    dtype: string
  - name: answer
    dtype: string
  splits:
  - name: train
    num_bytes: 25703144
    num_examples: 10000
  - name: test
    num_bytes: 2577392
    num_examples: 1000
  download_size: 12814944
  dataset_size: 28280536
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---