When tuning hyperparameters, I used k-fold data partition strategy and monitored the validation set accuracy as the metric to indicate the best performing set of hyperparame