Function std::module_evaluate

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

generic<M: Module> float module_evaluate(
    &M m,
    int loss,
    const &Dataset data,
    unsigned long batchSize,
)

Mean built-in loss over a dataset, without gradients.