►
Keras 2 API 文档 /
优化器
优化器
- 随机梯度下降（SGD）
- RMSprop
- 亚当优化器
- AdamW是一种优化算法，其结合了Adam优化器与权重衰减（Weight Decay）的技巧。这一方法有助于防止模型过拟合，从而提升模型的泛化能力。AdamW在Adam优化器的基础上引入了权重衰减，使得权重更新公式中包含了一个逐步减少权重值的项，从而有效地控制模型的复杂度。
- Adadelta
- Adagrad
- Adamax优化器
- Adafactor
- Nadam
- Ftrl（遵循正则化领导者）
apply_gradients method
Optimizer.apply_gradients(
grads_and_vars, name=None, skip_gradients_aggregation=False, **kwargs
)
Apply gradients to variables.
Arguments
- grads_and_vars: List of
(gradient, variable)pairs. - name: string, defaults to None. The name of the namescope to
use when creating variables. If None,
self.namewill be used. - skip_gradients_aggregation: If true, gradients aggregation will not be performed inside optimizer. Usually this arg is set to True when you write custom code aggregating gradients outside the optimizer.
- **kwargs: keyword arguments only used for backward compatibility.
Returns
A tf.Variable, representing the current iteration.
Raises
- TypeError: If
grads_and_varsis malformed. - RuntimeError: If called in a cross-replica context.
variables property
tf_keras.optimizers.Optimizer.variables
Returns variables of this optimizer.