certen/Sudoku-1M-Puzzles

---
license: mit
task_categories:
- tabular-classification
- reinforcement-learning
tags:
- sudoku
- puzzles
- logic
- synthetic
pretty_name: 1M Sudoku Puzzles (9x9)
size_categories:
- 100K<n<1M
configs:
- config_name: default
data_files:
- sudoku_dataset_1m.csv
---

# 1M Sudoku Puzzles (9x9)

A dataset of 1,000,000 unique 9x9 Sudoku puzzles and their corresponding solutions. Each puzzle has between 40 and 60 blank cells, spanning various difficulty levels.

## Dataset Structure

The dataset is provided in a single CSV file with the following columns:

- `puzzle`: 81-character string representing the initial board (0 = blank).
- `solution`: 81-character string representing the completed board.

### Example Row
```csv
puzzle,solution
00000001040000000000000000... , 526837941413925768...
```

## Data Generation
Puzzles were generated using a custom geometric transformation engine based on a seed board and then blanked out randomly. This method ensures all solutions are valid Sudoku boards.

## Usage
This dataset is designed for training and benchmarking machine learning models (Transformers, CNNs, GNNs) on Sudoku solving tasks.

### Loading with `datasets` library
```python
from datasets import load_dataset

dataset = load_dataset("certen/Sudoku-1M-Puzzles")
print(dataset["train"][0])
```

## License
MIT