README.md

---
configs:
  - config_name: calcudoko
    data_files:
      - split: train
        path: calcudoko/train-*.parquet
      - split: dev
        path: calcudoko/dev-*.parquet
      - split: test
        path: calcudoko/test-*.parquet
  - config_name: futoshiki
    data_files:
      - split: train
        path: futoshiki/train-*.parquet
      - split: dev
        path: futoshiki/dev-*.parquet
      - split: test
        path: futoshiki/test-*.parquet
  - config_name: math_path
    data_files:
      - split: train
        path: math_path/train-*.parquet
      - split: dev
        path: math_path/dev-*.parquet
      - split: test
        path: math_path/test-*.parquet
  - config_name: number_wall
    data_files:
      - split: train
        path: number_wall/train-*.parquet
      - split: dev
        path: number_wall/dev-*.parquet
      - split: test
        path: number_wall/test-*.parquet
  - config_name: operation
    data_files:
      - split: train
        path: operation/train-*.parquet
      - split: dev
        path: operation/dev-*.parquet
      - split: test
        path: operation/test-*.parquet
  - config_name: space_reasoning
    data_files:
      - split: train
        path: space_reasoning/train-*.parquet
      - split: dev
        path: space_reasoning/dev-*.parquet
      - split: test
        path: space_reasoning/test-*.parquet
  - config_name: sudoku
    data_files:
      - split: train
        path: sudoku/train-*.parquet
      - split: dev
        path: sudoku/dev-*.parquet
      - split: test
        path: sudoku/test-*.parquet
  - config_name: sudoku_easy
    data_files:
      - split: train
        path: sudoku_easy/train-*.parquet
      - split: dev
        path: sudoku_easy/dev-*.parquet
      - split: test
        path: sudoku_easy/test-*.parquet
  - config_name: survo
    data_files:
      - split: train
        path: survo/train-*.parquet
      - split: dev
        path: survo/dev-*.parquet
      - split: test
        path: survo/test-*.parquet
---

# SynLogic Dataset

This dataset contains multiple tasks for synthetic logic reasoning.

## Available Tasks

- calcudoko
- futoshiki
- math_path
- number_wall
- operation
- space_reasoning
- sudoku
- sudoku_easy
- survo

## Usage

```python
from datasets import load_dataset

# Load a specific task
ds = load_dataset("SeanWang0027/synlogic", "calcudoko")
```