Function std::trainer_accuracy

#include <std/ml/trainer.h> source

generic<T> float trainer_accuracy(
    &T model,
    fn &FloatTensor(&T, const &FloatTensor) logitsFn,
    const &Dataset data,
    unsigned long batchSize,
)

Classification accuracy over a dataset: 'logitsFn' maps a batch to [rows, classes] logits.