README.md

---
dataset_info:
  features:
  - name: filename
    dtype: string
  - name: pages
    list:
    - name: image_desc
      dtype: string
    - name: page_num
      dtype: int64
    - name: pdf_page
      dtype: int64
    - name: text
      dtype: string
    - name: words
      list:
      - name: height
        dtype: int64
      - name: left
        dtype: int64
      - name: text
        dtype: string
      - name: top
        dtype: int64
      - name: width
        dtype: int64
  - name: url
    dtype: string
  - name: embedding
    sequence: float64
  - name: processed_at
    dtype: string
  - name: pdf_page_offset
    dtype: int64
  splits:
  - name: train
    num_bytes: 4904377
    num_examples: 36
  download_size: 1503049
  dataset_size: 4904377
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---