softmax及优化

https://zhuanlan.zhihu.com/p/168562182
这个更清楚些
https://blog.csdn.net/bitcarmanlee/article/details/82320853

def softmax(x):
    max_x = np.max(x, axis=-1, keepdims=True)
    x = x - max_x

    ex = np.exp(x)
    sum_ex = np.sum(ex, axis=1, keepdims=True)

    result = ex / sum_ex

    result = np.clip(result, 1e-10, 1e10)
    return result