sasicodes commited on
Commit
adb5a2b
·
verified ·
1 Parent(s): 3e51500

Delete README.md

Browse files
Files changed (1) hide show
  1. README.md +0 -73
README.md DELETED
@@ -1,73 +0,0 @@
1
- ---
2
- license: mit
3
- dataset_info:
4
- features:
5
- - name: image
6
- dtype: image
7
- - name: text
8
- dtype: string
9
- splits:
10
- - name: train_shard_034
11
- num_bytes: 3070462057.0
12
- num_examples: 5000
13
- download_size: 3070433781
14
- dataset_size: 3070462057.0
15
- pretty_name: tamily 1
16
- language:
17
- - ta
18
- source_datasets:
19
- - sasicodes/solvari-1
20
- task_categories:
21
- - image-to-text
22
- tags:
23
- - Vaṭṭeḻuttu
24
- configs:
25
- - config_name: default
26
- data_files:
27
- - split: train_shard_034
28
- path: data/train_shard_034-*
29
- ---
30
-
31
-
32
- # Tamily-1: Ancient Tamil OCR Synthetic Dataset
33
-
34
- ## Description
35
-
36
- - **Repository:** [sasicodes/tamily-1](https://huggingface.co/datasets/sasicodes/tamily-1)
37
- - **Point of Contact:** [@sasicodes](https://huggingface.co/sasicodes)
38
-
39
- ### Summary
40
-
41
- Tamily-1 is an ancient Tamil OCR synthetic dataset generated from the first 200,000 rows of [Solvari-1](https://huggingface.co/datasets/sasicodes/solvari-1), a large Tamil text corpus. The dataset contains rendered images of Tamil text with various augmentations and styles, making it suitable for training OCR models.
42
-
43
- ### Fields
44
-
45
- - `image`: PNG image of rendered Tamil text
46
- - `text`: Original Tamil text
47
-
48
- ### Data Splits
49
-
50
- The dataset is split into shards of 5,000 samples each, named as `train_shard_XXX`.
51
-
52
- #### Annotation process
53
-
54
- Each text is rendered with:
55
- - Random paper style (Palm Leaf, Pale Palm Leaf, Red Stone, White Stone, Paper)
56
- - Random background style (No Lines, With Lines, Blurred, With Lines and Noise)
57
- - Random augmentation (Rotation, Perspective, Stain, Ink Bleed)
58
-
59
- ### License
60
-
61
- MIT License
62
-
63
-
64
- ```bibtex
65
- @misc{tamily-1,
66
- author = {sasicodes},
67
- title = {Tamily-1: Ancient Tamil OCR Synthetic Dataset},
68
- year = {2025},
69
- publisher = {Hugging Face},
70
- journal = {Hugging Face Hub},
71
- howpublished = {\url{https://huggingface.co/datasets/sasicodes/tamily-1}}
72
- }
73
- ```