Function _Relu

express/NeuralNetWorkOp.cpp:433–440

Given an input value x, it computes the output as x if x > 0 and slope * x if x <= 0. Args: x: A variable. slope: A float, a positive float value, it leakes the negative part by multiplying with `slope` rather than setting it to 0.0f. Returns: output: A variable with the same type as `x`. */

output: A variable with the same type as `x`.
*/
VARP _Relu(VARP x, float slope) {
    std::unique_ptr<OpT> relu(new OpT);
    relu->type                 = OpType_ReLU;
    relu->main.type            = OpParameter_Relu;
    relu->main.value           = new ReluT;
    relu->main.AsRelu()->slope = slope;
    return (Variable::create(Expr::create(relu.get(), {x})));
}
/*Given an input value x, it computes Rectified Linear 6: min(max(x, 0), 6).
Args:
x: A variable.