README.md

---
dataset_info:
  config_name: default
  features:
  - name: image
    dtype:
      image:
        decode: false
  - name: xml_content
    dtype: string
  - name: filename
    dtype: string
  - name: project_name
    dtype: string
  splits:
  - name: train
    num_examples: 58
    num_bytes: 619623508
  download_size: 619623508
  dataset_size: 619623508
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train/**/*.parquet
tags:
  - image-to-text
  - htr
  - trocr
  - transcription
  - pagexml
license: mit
---

# Dataset Card for missiven-rawxml

This dataset was created using pagexml-hf converter from Transkribus PageXML data.

## Dataset Summary

This dataset contains 58 samples across 1 split(s).

## Dataset Structure

### Data Splits

- **train**: 58 samples

### Dataset Size

- Approximate total size: 590.92 MB
- Total samples: 58

### Features

- **image**: `Image(mode=None, decode=False)`
- **xml_content**: `Value('string')`
- **filename**: `Value('string')`
- **project_name**: `Value('string')`

## Data Organization

Data is organized as parquet shards by split and project:
```
data/
├── <split>/
│   └── <project_name>/
│       └── <timestamp>-<shard>.parquet
```

The HuggingFace Hub automatically merges all parquet files when loading the dataset.

## Usage

```python
from datasets import load_dataset

# Load entire dataset
dataset = load_dataset("thodel/missiven-rawxml")

# Load specific split
train_dataset = load_dataset("thodel/missiven-rawxml", split="train")
```

### Projects Included

unknown_project