Optimizer that implements the Adamax algorithm.
It is a variant of Adam based on the infinity norm. Default parameters follow those provided in the paper. Adamax is sometimes superior to adam, specially in models with embeddings.
See Also
Constants
float | BETA_ONE_DEFAULT | |
float | BETA_TWO_DEFAULT | |
float | EPSILON_DEFAULT | |
String | FIRST_MOMENT | |
float | LEARNING_RATE_DEFAULT | |
String | SECOND_MOMENT |
Inherited Constants
Public Constructors
Public Methods
String |
getOptimizerName()
Get the Name of the optimizer.
|
Inherited Methods
Constants
public static final float BETA_ONE_DEFAULT
public static final float BETA_TWO_DEFAULT
public static final float EPSILON_DEFAULT
public static final String FIRST_MOMENT
public static final float LEARNING_RATE_DEFAULT
public static final String SECOND_MOMENT
Public Constructors
public Adamax (Graph graph)
Creates an Optimizer that implements the Adamax algorithm.
Parameters
graph | the TensorFlow graph |
---|
public Adamax (Graph graph, String name)
Creates an Optimizer that implements the Adamax algorithm.
Parameters
graph | the TensorFlow graph |
---|---|
name | name for the operations Created when applying gradients. Defaults to "Adamax". |
public Adamax (Graph graph, float learningRate)
Creates an Optimizer that implements the Adamax algorithm.
Parameters
graph | the TensorFlow graph |
---|---|
learningRate | The learning rate. |
public Adamax (Graph graph, String name, float learningRate)
Creates an Optimizer that implements the Adamax algorithm.
Parameters
graph | the TensorFlow graph |
---|---|
name | name for the operations Created when applying gradients. Defaults to "Adamax". |
learningRate | The learning rate. |
public Adamax (Graph graph, float learningRate, float betaOne, float betaTwo, float epsilon)
Creates an Optimizer that implements the Adamax algorithm.
Parameters
graph | the TensorFlow graph |
---|---|
learningRate | The learning rate. |
betaOne | The exponential decay rate for the 1st moment estimates. |
betaTwo | The exponential decay rate for the exponentially weighted infinity norm. |
epsilon | A small constant for numerical stability. |
public Adamax (Graph graph, String name, float learningRate, float betaOne, float betaTwo, float epsilon)
Creates an Optimizer that implements the Adamax algorithm.
Parameters
graph | the TensorFlow graph |
---|---|
name | name for the operations Created when applying gradients. Defaults to "Adamax". |
learningRate | The learning rate. |
betaOne | The exponential decay rate for the 1st moment estimates. |
betaTwo | The exponential decay rate for the exponentially weighted infinity norm. |
epsilon | A small constant for numerical stability. |
Public Methods
public String getOptimizerName ()
Get the Name of the optimizer.
Returns
- The optimizer name.