qkeras.qrecurrent

Quantized recurrent layers for Keras 3 / qkerasV3.

Classes

`QBidirectional`(args, *kwargs)	Quantized bidirectional wrapper.
`QGRU`(args, *kwargs)	Quantized GRU layer.
`QGRUCell`(args, *kwargs)	Quantized GRU cell.
`QLSTM`(args, *kwargs)	Quantized LSTM layer.
`QLSTMCell`(args, *kwargs)	Quantized LSTM cell.
`QSimpleRNN`(args, *kwargs)	Quantized SimpleRNN layer.
`QSimpleRNNCell`(args, *kwargs)	Quantized SimpleRNN cell.

class qkeras.qrecurrent.QBidirectional(*args, **kwargs)[source]

Bases: Bidirectional

Quantized bidirectional wrapper.

property activation

get_quantization_config()[source]

get_quantizers()[source]

class qkeras.qrecurrent.QGRU(*args, **kwargs)[source]

Bases: RNN

Quantized GRU layer.

property activation

property bias_constraint

property bias_initializer

property bias_quantizer

property bias_quantizer_internal

property bias_regularizer

call(sequences, initial_state=None, mask=None, training=False)[source]

compute_output_shape(inputs_shape)[source]

property dropout

classmethod from_config(config)[source]

Creates an operation from its config.

This method is the reverse of get_config, capable of instantiating the same operation from the config dictionary.

Note: If you override this method, you might receive a serialized dtype config, which is a dict. You can deserialize it as follows:

```python if “dtype” in config and isinstance(config[“dtype”], dict):

policy = dtype_policies.deserialize(config[“dtype”])

```

Parameters:: config – A Python dictionary, typically the output of get_config.
Returns:: An operation instance.

get_config()[source]

Returns the config of the object.

An object config is a Python dictionary (serializable) containing the information needed to re-instantiate it.

get_prunable_weights()[source]

get_quantization_config()[source]

get_quantizers()[source]

property implementation

property kernel_constraint

property kernel_initializer

property kernel_quantizer

property kernel_quantizer_internal

property kernel_regularizer

property recurrent_activation

property recurrent_constraint

property recurrent_dropout

property recurrent_initializer

property recurrent_quantizer

property recurrent_quantizer_internal

property recurrent_regularizer

property reset_after

property state_quantizer

property state_quantizer_internal

property units

property use_bias

class qkeras.qrecurrent.QGRUCell(*args, **kwargs)[source]

Bases: GRUCell

Quantized GRU cell.

call(inputs, states, training=False)[source]

get_config()[source]

Returns the config of the object.

An object config is a Python dictionary (serializable) containing the information needed to re-instantiate it.

class qkeras.qrecurrent.QLSTM(*args, **kwargs)[source]

Bases: RNN

Quantized LSTM layer.

property activation

property bias_constraint

property bias_initializer

property bias_quantizer

property bias_quantizer_internal

property bias_regularizer

call(sequences, initial_state=None, mask=None, training=False)[source]

compute_output_shape(inputs_shape)[source]

property dropout

classmethod from_config(config)[source]

Creates an operation from its config.

This method is the reverse of get_config, capable of instantiating the same operation from the config dictionary.

Note: If you override this method, you might receive a serialized dtype config, which is a dict. You can deserialize it as follows:

```python if “dtype” in config and isinstance(config[“dtype”], dict):

policy = dtype_policies.deserialize(config[“dtype”])

```

Parameters:: config – A Python dictionary, typically the output of get_config.
Returns:: An operation instance.

get_config()[source]

Returns the config of the object.

An object config is a Python dictionary (serializable) containing the information needed to re-instantiate it.

get_prunable_weights()[source]

get_quantization_config()[source]

get_quantizers()[source]

property implementation

property kernel_constraint

property kernel_initializer

property kernel_quantizer

property kernel_quantizer_internal

property kernel_regularizer

property recurrent_activation

property recurrent_constraint

property recurrent_dropout

property recurrent_initializer

property recurrent_quantizer

property recurrent_quantizer_internal

property recurrent_regularizer

property state_quantizer

property state_quantizer_internal

property unit_forget_bias

property units

property use_bias

class qkeras.qrecurrent.QLSTMCell(*args, **kwargs)[source]

Bases: LSTMCell

Quantized LSTM cell.

call(inputs, states, training=False)[source]

get_config()[source]

Returns the config of the object.

An object config is a Python dictionary (serializable) containing the information needed to re-instantiate it.

class qkeras.qrecurrent.QSimpleRNN(*args, **kwargs)[source]

Bases: RNN

Quantized SimpleRNN layer.

property activation

property bias_constraint

property bias_initializer

property bias_quantizer

property bias_quantizer_internal

property bias_regularizer

call(sequences, initial_state=None, mask=None, training=False)[source]

compute_output_shape(inputs_shape)[source]

property dropout

classmethod from_config(config)[source]

Creates an operation from its config.

This method is the reverse of get_config, capable of instantiating the same operation from the config dictionary.

Note: If you override this method, you might receive a serialized dtype config, which is a dict. You can deserialize it as follows:

```python if “dtype” in config and isinstance(config[“dtype”], dict):

policy = dtype_policies.deserialize(config[“dtype”])

```

Parameters:: config – A Python dictionary, typically the output of get_config.
Returns:: An operation instance.

get_config()[source]

Returns the config of the object.

An object config is a Python dictionary (serializable) containing the information needed to re-instantiate it.

get_prunable_weights()[source]

get_quantization_config()[source]

get_quantizers()[source]

property kernel_constraint

property kernel_initializer

property kernel_quantizer

property kernel_quantizer_internal

property kernel_regularizer

property recurrent_constraint

property recurrent_dropout

property recurrent_initializer

property recurrent_quantizer

property recurrent_quantizer_internal

property recurrent_regularizer

property state_quantizer

property state_quantizer_internal

property units

property use_bias

class qkeras.qrecurrent.QSimpleRNNCell(*args, **kwargs)[source]

Bases: SimpleRNNCell

Quantized SimpleRNN cell.

call(inputs, states, training=False)[source]

get_config()[source]

Returns the config of the object.

An object config is a Python dictionary (serializable) containing the information needed to re-instantiate it.