imone commited on
Commit
58942f9
1 Parent(s): 693d945

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +45 -0
README.md ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ task_categories:
3
+ - question-answering
4
+ ---
5
+ # Hardest Sudoku Puzzle Dataset V2
6
+
7
+ This dataset contains a mixture of easy and very hard Sudoku puzzles collected from the Sudoku community.
8
+
9
+ ## Dataset Composition
10
+
11
+ ### Sources
12
+
13
+ - [tdoku benchmarks](https://github.com/t-dillon/tdoku/blob/master/benchmarks/README.md#benchmarked-data-sets)
14
+ - [enjoysudoku](http://forum.enjoysudoku.com/the-hardest-sudokus-new-thread-t6539-600.html#p277835)
15
+
16
+ ### Easy Puzzles (1.1M)
17
+
18
+ - puzzles0_kaggle
19
+ - puzzles1_unbiased
20
+ - puzzles2_17_clue
21
+
22
+ ### Hard Puzzles (3.1M)
23
+
24
+ - puzzles3_magictour_top1465
25
+ - puzzles4_forum_hardest_1905
26
+ - puzzles6_forum_hardest_1106
27
+ - ph_2010/01_file1.txt
28
+
29
+ ## Dataset Characteristics
30
+
31
+ - All puzzles have been exact-deduped and randomly permuted by row, column, box, and digit.
32
+ - Each puzzle is guaranteed to have a unique solution.
33
+ - Puzzles in the train set are [mathematically inequivalent](http://sudopedia.enjoysudoku.com/Mathematically_equivalent.html) to those in the test set.
34
+
35
+ ## Dataset Structure
36
+
37
+ - Train set: `train.csv` (3.8M examples)
38
+ - Test set: `test.csv` (423k examples)
39
+
40
+ Puzzles and solutions are flattened in row-major order. Rating is evaluated by number of backtracks needed by [tdoku solver]((https://github.com/t-dillon/tdoku) required to solve the puzzle (higher is harder).
41
+
42
+ ## Usage Guidelines
43
+
44
+ 1. Train models using only the train set.
45
+ 2. Evaluate models on the test set using exact accuracy (all numbers must be correct).