Fast GRU implementation backed by cuDNN.
View aliases
Compat aliases for migration
See Migration guide for more details.
tf.keras.layers.CuDNNGRU(
units, kernel_initializer='glorot_uniform', recurrent_initializer='orthogonal',
bias_initializer='zeros', kernel_regularizer=None, recurrent_regularizer=None,
bias_regularizer=None, activity_regularizer=None, kernel_constraint=None,
recurrent_constraint=None, bias_constraint=None, return_sequences=False,
return_state=False, go_backwards=False, stateful=False, **kwargs
)
More information about cuDNN can be found on the NVIDIA developer website. Can only be run on GPU.
Arguments | |
---|---|
units
|
Positive integer, dimensionality of the output space. |
kernel_
|
Initializer for the kernel weights matrix, used for
the linear transformation of the inputs.
|
recurrent_
|
Initializer for the recurrent_ weights
matrix, used for the linear transformation of the recurrent state.
|
bias_
|
Initializer for the bias vector. |
kernel_
|
Regularizer function applied to the kernel weights
matrix.
|
recurrent_
|
Regularizer function applied to the
recurrent_ weights matrix.
|
bias_
|
Regularizer function applied to the bias vector. |
activity_
|
Regularizer function applied to the output of the layer (its "activation"). |
kernel_
|
Constraint function applied to the kernel weights
matrix.
|
recurrent_
|
Constraint function applied to the
recurrent_ weights matrix.
|
bias_
|
Constraint function applied to the bias vector. |
return_
|
Boolean. Whether to return the last output in the output sequence, or the full sequence. |
return_
|
Boolean. Whether to return the last state in addition to the output. |
go_
|
Boolean (default False). If True, process the input sequence backwards and return the reversed sequence. |
stateful
|
Boolean (default False). If True, the last state for each sample at index i in a batch will be used as initial state for the sample of index i in the following batch. |
Attributes | |
---|---|
cell
|
|
states
|
Methods
get_initial_state
get_initial_state(
inputs
)
reset_states
reset_states(
states=None
)