metadata
dataset_info:
features:
- name: source_id
dtype: string
- name: pdf_url
dtype: string
- name: status
dtype: string
- name: processed_at
dtype: string
- name: elapsed_seconds
dtype: float64
- name: num_pages
dtype: int64
- name: num_pages_processed
dtype: int64
- name: max_pages_per_paper
dtype: int64
- name: pdf_exceeds_page_limit
dtype: bool
- name: failed_pages
list: 'null'
- name: prompt_type
dtype: string
- name: model_id
dtype: string
- name: script_version
dtype: timestamp[ms]
- name: markdown
dtype: string
- name: html
dtype: string
- name: page_stats
list:
- name: page_number
dtype: int64
- name: error
dtype: bool
- name: token_count
dtype: int64
- name: num_chunks
dtype: int64
splits:
- name: train
num_bytes: 11935752
num_examples: 111
download_size: 3963785
dataset_size: 11935752
configs:
- config_name: default
data_files:
- split: train
path: data/train-*