Datasetpublic

starkdv123/mnist-numbers-0to10000-128x128

MNIST Numbers 0..10,000 (128×128) 10,000 synthetic grayscale images composed from MNIST digits (black on white), resized to 128×128. Each row corresponds to an integer n ∈ [0, 10,000] and includes: image: digits tiled left→right with small rotation jitter digits: e.g., "10000" words: e.g., "ten thousand" (no "and") value: integer 0..10,000 length: number of digits (1..5) Splits train: 9,000 test: 1,000 Usage from datasets import load_dataset DS… See the full description on the dataset page: https://huggingface.co/datasets/starkdv123/mnist-numbers-0to10000-128x128.

sourceHugging Facemitupdated 1y agoView on Hugging Face

0likes33downloads

Dataset Card

MNIST Numbers 0..10,000 (128×128)

10,000 synthetic grayscale images composed from MNIST digits (black on white), resized to 128×128. Each row corresponds to an integer n ∈ [0, 10,000] and includes:

—`image`: digits tiled left→right with small rotation jitter
—`digits`: e.g., "10000"
—`words`: e.g., "ten thousand" (no "and")
—`value`: integer 0..10,000
—`length`: number of digits (1..5)

Splits

—train: 9,000
—test: 1,000

Usage

python

from datasets import load_dataset
DS = load_dataset("starkdv123/mnist-numbers-0to10000-128x128")
ex = DS["train"][0]
ex["image"].show()
print(ex["value"], ex["digits"], " | ", ex["words"])

Notes

—Digits are sampled from torchvision.datasets.MNIST(train=True).
—Mild rotation jitter (±10°); composed horizontally then resized to square.
—Images are file-backed (PNG) to ensure Hub Dataset Viewer compatibility (Parquet auto-conversion).