baobabtech/test-eval-docs-docling-plain
收藏Hugging Face2025-12-02 更新2026-01-03 收录
下载链接：
https://hf-mirror.com/datasets/baobabtech/test-eval-docs-docling-plain
下载链接
链接失效反馈官方服务：
资源简介：
---
tags:
- document-processing
- docling
- hierarchical-parsing
- pdf-processing
- generated
---
# PDF Document Processing with Docling
This dataset contains structured markdown extraction from PDFs in [baobabtech/test-eval-documents](https://huggingface.co/datasets/baobabtech/test-eval-documents)
using Docling with hierarchical parsing.
## Processing Details
- **Source Dataset**: [baobabtech/test-eval-documents](https://huggingface.co/datasets/baobabtech/test-eval-documents)
- **Number of PDFs**: 20
- **Processing Time**: 8.4 minutes
- **Processing Date**: 2025-12-02 15:40 UTC
### Configuration
- **PDF Column**: `pdf_bytes`
- **Dataset Split**: `train`
## Dataset Structure
The dataset contains all original columns plus:
- `original_md`: Markdown extracted by Docling (before hierarchical restructuring)
- `hierarchical_md`: Markdown with proper heading hierarchy (after hierarchical processing)
- `sections_toc`: Table of contents (one section per line, indented by level)
- `inference_info`: JSON with processing metadata
## Usage
```python
from datasets import load_dataset
dataset = load_dataset("YOUR_DATASET_ID", split="train")
for example in dataset:
print(f"Document: {example.get('file_name', 'unknown')}")
# Original markdown from Docling
print("=== Original Markdown ===")
print(example['original_md'][:500])
# Hierarchical markdown with proper heading levels
print("\n=== Hierarchical Markdown ===")
print(example['hierarchical_md'][:500])
# Table of contents
print("\n=== Table of Contents ===")
print(example['sections_toc'])
break
```
提供机构：
baobabtech
