m53297601
/
Paddle
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

unify the bias

Browse Source
mobile_baidu
peterzhang2029 8 years ago
parent 2b93934d2a
commit e97354152c
2 changed files with 77 additions and 60 deletions
Show Stats Download Patch File Download Diff File

51
python/paddle/trainer_config_helpers/layers.py
Unescape View File

@ -3649,9 +3649,10 @@ def gru_step_layer(input,
:param name: The name of this layer. It is optional.
:param gate_act: Activation type of this layer's two gates. Default is Sigmoid.
:type gate_act: BaseActivation
:param bias_attr: The bias attribute. If the parameter is set to False or an object
whose type is not ParameterAttribute, no bias is defined. If the
parameter is set to True, the bias is initialized to zero.
:param bias_attr: The parameter attribute for bias. If this parameter is set to
False or an object whose type is not ParameterAttribute, no bias
is defined. If this parameter is set to True,
the bias is initialized to zero.
:type bias_attr: ParameterAttribute | None | bool | Any
:param param_attr: the parameter_attribute for transforming the output_mem
from previous step.
@ -3711,9 +3712,10 @@ def gru_step_naive_layer(input,
:type act: BaseActivation
:param gate_act: Activation type of this layer's two gates. Default is Sigmoid.
:type gate_act: BaseActivation
:param bias_attr: The bias attribute. If the parameter is set to False or an object
whose type is not ParameterAttribute, no bias is defined. If the
parameter is set to True, the bias is initialized to zero.
:param bias_attr: The parameter attribute for bias. If this parameter is set to
False or an object whose type is not ParameterAttribute, no bias
is defined. If this parameter is set to True,
the bias is initialized to zero.
:type bias_attr: ParameterAttribute | None | bool | Any
:param param_attr:
:param layer_attr:
@ -3843,9 +3845,10 @@ def recurrent_layer(input,
:type input: LayerOutput
:param act: Activation type. TanhActivation is the default.
:type act: BaseActivation
:param bias_attr: The bias attribute. If the parameter is set to False or an object
whose type is not ParameterAttribute, no bias is defined. If the
parameter is set to True, the bias is initialized to zero.
:param bias_attr: The parameter attribute for bias. If this parameter is set to
False or an object whose type is not ParameterAttribute,
no bias is defined. If the parameter is set to True,
the bias is initialized to zero.
:type bias_attr: ParameterAttribute | None | bool | Any
:param param_attr: parameter attribute.
:type param_attr: ParameterAttribute
@ -4835,9 +4838,10 @@ def tensor_layer(a,
:type act: BaseActivation
:param param_attr: The Parameter Attribute.
:type param_attr: ParameterAttribute
:param bias_attr: The bias attribute. If the parameter is set to False or an object
whose type is not ParameterAttribute, no bias is defined. If the
parameter is set to True, the bias is initialized to zero.
:param bias_attr: The parameter attribute for bias. If this parameter is set to
False or an object whose type is not ParameterAttribute,
no bias is defined. If this parameter is set to True,
the bias is initialized to zero.
:type bias_attr: ParameterAttribute | None | bool | Any
:param layer_attr: Extra Layer config.
:type layer_attr: ExtraLayerAttribute | None
@ -4899,9 +4903,10 @@ def selective_fc_layer(input,
:type act: BaseActivation
:param param_attr: The Parameter Attribute.
:type param_attr: ParameterAttribute
:param bias_attr: The bias attribute. If the parameter is set to False or an object
whose type is not ParameterAttribute, no bias is defined. If the
parameter is set to True, the bias is initialized to zero.
:param bias_attr: The parameter attribute for bias. If this parameter is set to
False or an object whose type is not ParameterAttribute,
no bias is defined. If this parameter is set to True,
the bias is initialized to zero.
:type bias_attr: ParameterAttribute | None | bool | Any
:param layer_attr: Extra Layer config.
:type layer_attr: ExtraLayerAttribute | None
@ -5584,10 +5589,10 @@ def nce_layer(input,
to the num_classes. Each member of the list defines
the probability of a class given input x.
:type neg_distribution: list | tuple | collections.Sequence | None
:param bias_attr: The attribute for bias. If this parameter is set False or
any object whose type is not ParameterAttribute, no bias
is added. If this parameter is set True, the bias is
initialized to zero.
:param bias_attr: The parameter attribute for bias. If this parameter is set to
False or an object whose type is not ParameterAttribute,
no bias is defined. If this parameter is set to True,
the bias is initialized to zero.
:type bias_attr: ParameterAttribute | None | bool | Any
:param layer_attr: Extra Layer Attribute.
:type layer_attr: ExtraLayerAttribute
@ -6497,9 +6502,9 @@ def gated_unit_layer(input,
:param gate_param_attr: The parameter attribute of the gate. See ParameterAttribute
for details.
:type gate_param_attr: ParameterAttribute
:param gate_bias_attr: The bias attribute of the gate. If the parameter is set to False or
:param gate_bias_attr: The bias attribute of the gate. If this parameter is set to False or
an object whose type is not ParameterAttribute, no bias is defined.
If the parameter is set to True, the bias is initialized to zero.
If this parameter is set to True, the bias is initialized to zero.
:type gate_bias_attr: ParameterAttribute | bool | None | Any
:param inproj_attr: Extra layer attributes of the projection. See ExtraLayerAttribute for
details.
@ -6507,9 +6512,9 @@ def gated_unit_layer(input,
:param inproj_param_attr: The parameter attribute of the projection. See ParameterAttribute
for details.
:type inproj_param_attr: ParameterAttribute
:param inproj_bias_attr: The bias attribute of the projection. If the parameter is set to False
:param inproj_bias_attr: The bias attribute of the projection. If this parameter is set to False
or an object whose type is not ParameterAttribute, no bias is defined.
If the parameter is set to True, the bias is initialized to zero.
If this parameter is set to True, the bias is initialized to zero.
:type inproj_bias_attr: ParameterAttribute | bool | None | Any
:param layer_attr: Extra layer attribute of the product. See ExtraLayerAttribute for
details.

86
python/paddle/trainer_config_helpers/networks.py
Unescape View File

@ -681,36 +681,42 @@ def lstmemory_unit(input,
state_act=TanhActivation())
:param input: input layer.
:param input: Input layer.
:type input: LayerOutput
:param out_memory: output of previous time step
:param out_memory: The output of previous time step.
:type out_memory: LayerOutput | None
:param name: lstmemory unit name.
:param name: The lstmemory unit name.
:type name: basestring
:param size: lstmemory unit size.
:param size: The lstmemory unit size.
:type size: int
:param param_attr: parameter attribute, None means default attribute.
:param param_attr: The parameter attribute for the weights in
input to hidden projection.
None means default attribute.
:type param_attr: ParameterAttribute
:param act: last activiation type of lstm.
:param act: The last activiation type of lstm.
:type act: BaseActivation
:param gate_act: gate activiation type of lstm.
:param gate_act: The gate activiation type of lstm.
:type gate_act: BaseActivation
:param state_act: state activiation type of lstm.
:param state_act: The state activiation type of lstm.
:type state_act: BaseActivation
:param input_proj_bias_attr: bias attribute for input to hidden projection.
False or None means no bias. If the parameter is set to True,
the bias is initialized to zero.
:param input_proj_bias_attr: The parameter attribute for the bias in
input to hidden projection.
False or None means no bias.
If this parameter is set to True,
the bias is initialized to zero.
:type input_proj_bias_attr: ParameterAttribute|bool|None
:param input_proj_layer_attr: extra layer attribute for input to hidden
projection of the LSTM unit, such as dropout, error clipping.
:param input_proj_layer_attr: The extra layer attribute for
input to hidden projection of the LSTM unit,
such as dropout, error clipping.
:type input_proj_layer_attr: ExtraLayerAttribute
:param lstm_bias_attr: bias parameter attribute of lstm layer.
If the parameter is set to True or None,
the bias is initialized to zero.
:param lstm_bias_attr: The parameter attribute for the bias in lstm layer.
False or None means no bias.
If this parameter is set to True,
the bias is initialized to zero.
:type lstm_bias_attr: ParameterAttribute|True|None
:param lstm_layer_attr: extra attribute of lstm layer.
:param lstm_layer_attr: The extra attribute of lstm layer.
:type lstm_layer_attr: ExtraLayerAttribute
:return: lstmemory unit name.
:return: The lstmemory unit name.
:rtype: LayerOutput
"""
if size is None:
@ -788,36 +794,42 @@ def lstmemory_group(input,
gate_act=SigmoidActivation(),
state_act=TanhActivation())
:param input: input layer.
:param input: Input layer.
:type input: LayerOutput
:param size: lstmemory group size.
:param size: The lstmemory group size.
:type size: int
:param name: name of lstmemory group.
:param name: The name of lstmemory group.
:type name: basestring
:param out_memory: output of previous time step.
:param out_memory: The output of previous time step.
:type out_memory: LayerOutput | None
:param reverse: process the input in a reverse order or not.
:param reverse: Process the input in a reverse order or not.
:type reverse: bool
:param param_attr: parameter attribute, None means default attribute.
:param param_attr: The parameter attribute for the weights in
input to hidden projection.
None means default attribute.
:type param_attr: ParameterAttribute
:param act: last activiation type of lstm.
:param act: The last activiation type of lstm.
:type act: BaseActivation
:param gate_act: gate activiation type of lstm.
:param gate_act: The gate activiation type of lstm.
:type gate_act: BaseActivation
:param state_act: state activiation type of lstm.
:param state_act: The state activiation type of lstm.
:type state_act: BaseActivation
:param lstm_bias_attr: bias parameter attribute of lstm layer.
If the parameter is set to True or None, the bias is
initialized to zero.
:type lstm_bias_attr: ParameterAttribute|True|None
:param input_proj_bias_attr: bias attribute for input to hidden projection.
False or None means no bias. If the parameter is set to True,
the bias is initialized to zero.
:param input_proj_bias_attr: The parameter attribute for the bias in
input to hidden projection.
False or None means no bias.
If this parameter is set to True,
the bias is initialized to zero.
:type input_proj_bias_attr: ParameterAttribute|bool|None
:param input_proj_layer_attr: extra layer attribute for input to hidden
projection of the LSTM unit, such as dropout, error clipping.
:param input_proj_layer_attr: The extra layer attribute for
input to hidden projection of the LSTM unit,
such as dropout, error clipping.
:type input_proj_layer_attr: ExtraLayerAttribute
:param lstm_layer_attr: lstm layer's extra attribute.
:param lstm_bias_attr: The parameter attribute for the bias in lstm layer.
False or None means no bias.
If this parameter is set to True,
the bias is initialized to zero.
:type lstm_bias_attr: ParameterAttribute|True|None
:param lstm_layer_attr: The extra attribute of lstm layer.
:type lstm_layer_attr: ExtraLayerAttribute
:return: the lstmemory group.
:rtype: LayerOutput

Write Preview
Loading…
Cancel
Save