std :: ml::nn
#include <std/ml/nn.h> source ยท definition
#include <std/ml/nn.h>
float tensor_accuracy(const &FloatTensor logits, const &FloatTensor labels)
Fraction of rows whose argmax equals the label.