Accuracy and Recall is the same for resnet implementation pytorch

Question

I have used accuracy_score and recall_score from sklearn.metrics to calculate the scores, but no matter what epoch it is, both the accuracy and recall are coming up as the same value for CIFAR100 dataset. I am trying to implement it in Pytorch. Sample y_pred, y_true that I am passing to the function look like y_pred = [11, 12, 65, ..... 13, 15, 17] and y_true = [11, 12, 53, ..... 13, 78, 17]. Can you suggest where I might be going wrong?