3 ms·
One big problem is the network is never trained on incorrect Sudokus, so at that point it's effectively into random guessing.
by CJefferson 9y ago
One big problem is the network is never trained on incorrect Sudokus, so at that point it's effectively into random guessing.