File size: 3,368 Bytes
011d983
 
37fa728
 
6bad4fc
37fa728
 
 
 
6bad4fc
37fa728
 
 
02412bd
6bad4fc
02412bd
 
 
f797227
 
 
 
 
011d983
37fa728
 
 
 
 
0669726
 
 
37fa728
 
 
f797227
 
37fa728
 
 
 
896ea6a
37fa728
 
 
 
 
02412bd
0669726
37fa728
 
 
205acff
 
37fa728
 
bbb13e2
 
 
 
 
37fa728
 
205acff
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f797227
 
37fa728
f797227
37fa728
 
205acff
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
---

license: agpl-3.0
configs:
- config_name: asciiart
  default: true
  data_files:
  - split: train
    path: "asciiart/train/*.parquet"
- config_name: copypasta
  default: false
  data_files:
  - split: train
    path: "copypasta/train/*.parquet"
- config_name: graffiti
  default: false
  data_files:
  - split: train
    path: "graffiti/train/*.parquet"
- config_name: images
  default: false
  data_files:
  - split: train
    path: "images/train/*.parquet"
---


# ASCII Art

## Description

This is a text-to-image dataset, where the images are actually ASCII art.

The ASCII arts come from various sources:

- `asciiart`: made by independent artists, listed on [asciiart.eu][source-asciiart]
- `copypasta`: common twitch emotes, listed on [twitchquotes.com][source-twitchquotes]
- `graffiti`: text samples styled using various ASCII art fonts with [a tool][github-graffiti-tool]
- `images`: conversion of a portion of the dataset [DataCompDR-12M][huggingface-dataset] using [a tool][github-conversion-tool]

## Metadata

- homepage: [https://github.com/apehex/scrapscii][github-scrapscii]
- version: 1.0.0

| Config            | Split     | Size      | Samples   |
| ----------------- | --------- | --------- | --------- |
| 'asciiart'        | 'train'   | 1.8 MB    | 5293      |
| 'copypasta'       | 'train'   | 488 KB    | 909       |
| 'graffiti'        | 'train'   | 25 MB     | 21504     |
| 'images'          | 'train'   | 315 MB    | 40960     |

## Features

All the features of the dataset are textual:

```python

datasets.Features({

    'caption': datasets.features.Value(dtype='string'),

    'content': datasets.features.Value(dtype='string'),

    'labels': datasets.features.Value(dtype='string'),

    'charsets': datasets.features.Value(dtype='string'),

    'chartypes': datasets.features.Value(dtype='string'),})

```

- `caption`:
    - a brief description of the "image"
    - for example `"Undertale Sans full body"`
    - can be empty
- `content`:
    - the actual ASCII art, in UTF-8 encoding
    - for example `"(╯°□°)╯︵ ┻━┻ FLIP THAT TABLE.\r\n┻━┻ ︵ ヽ(°□°ヽ) FLIP THIS TABLE."`
    - is never empty
- `labels`:
    - a list of tags, separated by commas
    - the set of tags depends on the dataset
    - for example `"Cartoons,Smurfs"`
    - can be empty
- `charsets`:
    - the [Unicode sections][unicode-nameslist] of the characters in the AA
    - for example `"Braille Patterns,ASCII C0"`
    - "C0 Controls and Basic Latin" is renamed "ASCII C0"
    - "C1 Controls and Latin-1 Supplement" is renamed "ASCII C1"
    - is never empty
- `chartypes`:
    - the [types of characters][wiki-category] in the AA
    - for example `"Punctuations,Others,Symbols,Letters"`
    - is never empty

[github-conversion-tool]: https://github.com/TheZoraiz/ascii-image-converter
[github-graffiti-tool]: https://github.com/sepandhaghighi/art
[github-scrapscii]: https://github.com/apehex/scrapscii/
[huggingface-dataset]: https://huggingface.co/datasets/apple/DataCompDR-12M
[source-asciiart]: https://www.asciiart.eu/
[source-twitchquotes]: https://www.twitchquotes.com/
[unicode-nameslist]: https://www.unicode.org/charts/nameslist/
[wiki-category]: https://en.wikipedia.org/wiki/Unicode_character_property#General_Category