zhayushui的专栏

文章目录介绍softmaxsoftmaxloss介绍softmax家族总结softmaximport numpy as npdef softmax(X,SUBMAX=False): """dtype: float X shape: batch * -1 * classes ndarray or : -1 * classes nd...

1.ndarray数据交换指定的两个维度比如：矩阵A=1x3x2 => A=3x1x2A=np.swapaxes(A,0,1)参考：https://docs.scipy.org/doc/numpy-1.13.0/reference/generated/numpy.swapaxes.html