Tensorflow:Adam Optimizer使用

tf.train.AdamOptimizer官网接口文档

 def __init__(self, learning_rate=0.001, beta1=0.9, beta2=0.999, epsilon=1e-8,
               use_locking=False, name="Adam"):

公式原型

Args:

  • learning_rate: A Tensor or a floating point value. The learning rate.
    学习率
  • beta1: A float value or a constant float tensor. The exponential decay rate for the 1st moment estimates.
    指数衰减率的一阶矩估计
  • beta2: A float value or a constant float tensor. The exponential decay rate for the 2nd moment estimates.
    指数衰减率的二阶矩估计
  • epsilon: A small constant for numerical stability.
    防止分母除以0
    论文截图:
    算法细节

进一步学习可以看文章:
你真的懂Adam吗?

你可能感兴趣的:(Tensorflow:Adam Optimizer使用)