Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,92 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
{}
|
3 |
+
---
|
4 |
+
|
5 |
+
# Dataset Card for KMNIST
|
6 |
+
|
7 |
+
<!-- Provide a quick summary of the dataset. -->
|
8 |
+
|
9 |
+
## Dataset Details
|
10 |
+
|
11 |
+
### Dataset Description
|
12 |
+
|
13 |
+
<!-- Provide a longer summary of what this dataset is. -->
|
14 |
+
This dataset contains two variants, **Kuzushiji-MNIST** and **Kuzushiji-49**.
|
15 |
+
|
16 |
+
**Kuzushiji-MNIST** is a drop-in replacement for the MNIST dataset.
|
17 |
+
|
18 |
+
**Kuzushiji-49**, as the name suggests, has 49 classes, is a much larger, but imbalanced dataset containing 48 Hiragana characters and one Hiragana iteration mark.
|
19 |
+
|
20 |
+
|
21 |
+
### Dataset Sources
|
22 |
+
|
23 |
+
<!-- Provide the basic links for the dataset. -->
|
24 |
+
|
25 |
+
- **Homepage:** https://github.com/rois-codh/kmnist
|
26 |
+
- **Paper:** Clanuwat, T., Bober-Irizar, M., Kitamoto, A., Lamb, A., Yamamoto, K., & Ha, D. (2018). Deep learning for classical japanese literature. arXiv preprint arXiv:1812.01718.
|
27 |
+
|
28 |
+
## Dataset Structure
|
29 |
+
|
30 |
+
<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
|
31 |
+
|
32 |
+
#### Kuzushiji-MNIST:
|
33 |
+
|
34 |
+
Total images: 70,000
|
35 |
+
|
36 |
+
Classes: 10 categories
|
37 |
+
|
38 |
+
Splits:
|
39 |
+
|
40 |
+
- **Train:** 60,000 images
|
41 |
+
|
42 |
+
- **Test:** 10,000 images
|
43 |
+
|
44 |
+
Image specs: 28×28 pixels, grayscale
|
45 |
+
|
46 |
+
#### Kuzushiji-49:
|
47 |
+
|
48 |
+
Total images: 270,912
|
49 |
+
|
50 |
+
Classes: 49 categories
|
51 |
+
|
52 |
+
Splits:
|
53 |
+
|
54 |
+
- **Train:** 232,365 images
|
55 |
+
|
56 |
+
- **Test:** 38,547 images
|
57 |
+
|
58 |
+
Image specs: 28×28 pixels, grayscale
|
59 |
+
|
60 |
+
|
61 |
+
## Example Usage
|
62 |
+
Below is a quick example of how to load this dataset via the Hugging Face Datasets library.
|
63 |
+
```
|
64 |
+
from datasets import load_dataset
|
65 |
+
|
66 |
+
# Load the dataset
|
67 |
+
dataset = load_dataset("randall-lab/kmnist", name="kmnist", split="train", trust_remote_code=True)
|
68 |
+
# dataset = load_dataset("randall-lab/kmnist", name="kmnist", split="test", trust_remote_code=True)
|
69 |
+
# dataset = load_dataset("randall-lab/kmnist", name="k49mnist", split="train", trust_remote_code=True)
|
70 |
+
# dataset = load_dataset("randall-lab/kmnist", name="k49mnist", split="test", trust_remote_code=True)
|
71 |
+
|
72 |
+
# Access a sample from the dataset
|
73 |
+
example = dataset[0]
|
74 |
+
image = example["image"]
|
75 |
+
label = example["label"]
|
76 |
+
|
77 |
+
image.show() # Display the image
|
78 |
+
print(f"Label: {label}")
|
79 |
+
```
|
80 |
+
|
81 |
+
## Citation
|
82 |
+
|
83 |
+
<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
|
84 |
+
|
85 |
+
**BibTeX:**
|
86 |
+
|
87 |
+
@article{clanuwat2018deep,
|
88 |
+
title={Deep learning for classical japanese literature},
|
89 |
+
author={Clanuwat, Tarin and Bober-Irizar, Mikel and Kitamoto, Asanobu and Lamb, Alex and Yamamoto, Kazuaki and Ha, David},
|
90 |
+
journal={arXiv preprint arXiv:1812.01718},
|
91 |
+
year={2018}
|
92 |
+
}
|