Created by: ceci3

Add bce_loss op, paddle.nn.loss.BCELoss(weight=None, reduction='mean').

base format is Out = -1 * (label * log(input) + (1 - label) * log(1 - input))

examples:

            import paddle.fluid as fluid
            import numpy as np
            input = fluid.data(name="input", shape=[3, 1])
            label = fluid.data(name="label", shape=[3, 1])
            output = fluid.layers.bce_loss(input,label)
            place = fluid.CPUPlace()
            exe = fluid.Executor(place)
            exe.run(fluid.default_startup_program())

            input_data = np.array([0.5, 0.6, 0.7]).astype("float32")
            label_data = np.array([1.0, 0.0, 1.0]).astype("float32")
            output_data = exe.run(fluid.default_main_program(),
                    feed={"input":input_data, "label":label_data},
                    fetch_list=[output],
                    return_numpy=True)

            print(output_data)  # [array([0.65537095], dtype=float32)]