Update dataset card
Browse files
README.md
ADDED
|
@@ -0,0 +1,100 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
dataset_info:
|
| 3 |
+
config_name: default
|
| 4 |
+
features:
|
| 5 |
+
- name: image
|
| 6 |
+
dtype:
|
| 7 |
+
image:
|
| 8 |
+
decode: false
|
| 9 |
+
- name: text
|
| 10 |
+
dtype: string
|
| 11 |
+
- name: window_size
|
| 12 |
+
dtype: int64
|
| 13 |
+
- name: window_index
|
| 14 |
+
dtype: int64
|
| 15 |
+
- name: line_ids
|
| 16 |
+
dtype: string
|
| 17 |
+
- name: line_reading_order
|
| 18 |
+
dtype: string
|
| 19 |
+
- name: filename
|
| 20 |
+
dtype: string
|
| 21 |
+
- name: project_name
|
| 22 |
+
dtype: string
|
| 23 |
+
splits:
|
| 24 |
+
- name: train
|
| 25 |
+
num_examples: 82
|
| 26 |
+
num_bytes: 344898560
|
| 27 |
+
download_size: 344898560
|
| 28 |
+
dataset_size: 344898560
|
| 29 |
+
configs:
|
| 30 |
+
- config_name: default
|
| 31 |
+
data_files:
|
| 32 |
+
- split: train
|
| 33 |
+
path: data/train/**/*.parquet
|
| 34 |
+
tags:
|
| 35 |
+
- image-to-text
|
| 36 |
+
- htr
|
| 37 |
+
- trocr
|
| 38 |
+
- transcription
|
| 39 |
+
- pagexml
|
| 40 |
+
license: mit
|
| 41 |
+
---
|
| 42 |
+
|
| 43 |
+
# Dataset Card for window-test-cli
|
| 44 |
+
|
| 45 |
+
This dataset was created using pagexml-hf converter from Transkribus PageXML data.
|
| 46 |
+
|
| 47 |
+
## Dataset Summary
|
| 48 |
+
|
| 49 |
+
This dataset contains 82 samples across 1 split(s).
|
| 50 |
+
|
| 51 |
+
## Dataset Structure
|
| 52 |
+
|
| 53 |
+
### Data Splits
|
| 54 |
+
|
| 55 |
+
- **train**: 82 samples
|
| 56 |
+
|
| 57 |
+
### Dataset Size
|
| 58 |
+
|
| 59 |
+
- Approximate total size: 328.92 MB
|
| 60 |
+
- Total samples: 82
|
| 61 |
+
|
| 62 |
+
### Features
|
| 63 |
+
|
| 64 |
+
- **image**: `Image(mode=None, decode=False)`
|
| 65 |
+
- **text**: `Value('string')`
|
| 66 |
+
- **window_size**: `Value('int64')`
|
| 67 |
+
- **window_index**: `Value('int64')`
|
| 68 |
+
- **line_ids**: `Value('string')`
|
| 69 |
+
- **line_reading_order**: `Value('string')`
|
| 70 |
+
- **filename**: `Value('string')`
|
| 71 |
+
- **project_name**: `Value('string')`
|
| 72 |
+
|
| 73 |
+
## Data Organization
|
| 74 |
+
|
| 75 |
+
Data is organized as parquet shards by split and project:
|
| 76 |
+
```
|
| 77 |
+
data/
|
| 78 |
+
├── <split>/
|
| 79 |
+
│ └── <project_name>/
|
| 80 |
+
│ └── <timestamp>-<shard>.parquet
|
| 81 |
+
```
|
| 82 |
+
|
| 83 |
+
The HuggingFace Hub automatically merges all parquet files when loading the dataset.
|
| 84 |
+
|
| 85 |
+
## Usage
|
| 86 |
+
|
| 87 |
+
```python
|
| 88 |
+
from datasets import load_dataset
|
| 89 |
+
|
| 90 |
+
# Load entire dataset
|
| 91 |
+
dataset = load_dataset("jwidmer/window-test-cli")
|
| 92 |
+
|
| 93 |
+
# Load specific split
|
| 94 |
+
train_dataset = load_dataset("jwidmer/window-test-cli", split="train")
|
| 95 |
+
```
|
| 96 |
+
|
| 97 |
+
### Projects Included
|
| 98 |
+
|
| 99 |
+
1505-02-10_Hanserezess,_Lübeck_Dienstag_nach_Scholastice_1505_(SAHST_Rep__2,_I_040-4)
|
| 100 |
+
|