o i)@sddlZddlmZddlZddlmZddlmZmZddlm Z ddlm Z dd l m Z dd lmZgd ZGd d d e ZGddde ZGdddeZGddde ZGdddeeZdS)N)Any)Tensor) ParameterUninitializedParameter) functional)init)Module)LazyModuleMixin)BilinearIdentity LazyLinearLinearcs@eZdZdZdededdffdd Zdedefd d ZZS) r aA placeholder identity operator that is argument-insensitive. Args: args: any argument (unused) kwargs: any keyword argument (unused) Shape: - Input: :math:`(*)`, where :math:`*` means any number of dimensions. - Output: :math:`(*)`, same shape as the input. Examples:: >>> m = nn.Identity(54, unused_argument1=0.1, unused_argument2=False) >>> input = torch.randn(128, 20) >>> output = m(input) >>> print(output.size()) torch.Size([128, 20]) argskwargsreturnNcstdSNsuper__init__)selfrr __class__f/var/www/html/eduruby.in/lip-sync/lip-sync-env/lib/python3.10/site-packages/torch/nn/modules/linear.pyr)szIdentity.__init__inputcCs|Srrrrrrrforward,szIdentity.forward) __name__ __module__ __qualname____doc__rrrr __classcell__rrrrr sr c seZdZUdZddgZeed<eed<eed<  ddedededdffd d Z dd d Z d edefddZ de fddZ ZS)ra&Applies a linear transformation to the incoming data: :math:`y = xA^T + b` This module supports :ref:`TensorFloat32`. On certain ROCm devices, when using float16 inputs this module will use :ref:`different precision` for backward. Args: in_features: size of each input sample out_features: size of each output sample bias: If set to ``False``, the layer will not learn an additive bias. Default: ``True`` Shape: - Input: :math:`(*, H_{in})` where :math:`*` means any number of dimensions including none and :math:`H_{in} = \text{in\_features}`. - Output: :math:`(*, H_{out})` where all but the last dimension are the same shape as the input and :math:`H_{out} = \text{out\_features}`. Attributes: weight: the learnable weights of the module of shape :math:`(\text{out\_features}, \text{in\_features})`. The values are initialized from :math:`\mathcal{U}(-\sqrt{k}, \sqrt{k})`, where :math:`k = \frac{1}{\text{in\_features}}` bias: the learnable bias of the module of shape :math:`(\text{out\_features})`. If :attr:`bias` is ``True``, the values are initialized from :math:`\mathcal{U}(-\sqrt{k}, \sqrt{k})` where :math:`k = \frac{1}{\text{in\_features}}` Examples:: >>> m = nn.Linear(20, 30) >>> input = torch.randn(128, 20) >>> output = m(input) >>> print(output.size()) torch.Size([128, 30]) in_features out_featuresweightTNbiasrcsr||d}t||_||_ttj||ffi||_|r-ttj|fi||_n| dd| dSNdevicedtyper') rrr$r%rtorchemptyr&r'register_parameterreset_parameters)rr$r%r'r*r+factory_kwargsrrrrZs    zLinear.__init__cCsdtj|jtdd|jdur0t|j\}}|dkr#dt|nd}t|j| |dSdS)N)arr )rZkaiming_uniform_r&mathsqrtr'Z_calculate_fan_in_and_fan_outuniform_)rZfan_in_boundrrrr/gs  zLinear.reset_parametersrcCst||j|jSr)FZlinearr&r'rrrrrqszLinear.forwardcCs d|jd|jd|jduS)Nz in_features=z, out_features=z, bias=)r$r%r'rrrr extra_reprts zLinear.extra_reprTNNrNrr r!r"Z __constants__int__annotations__rboolrr/rstrr:r#rrrrr0s $  rc s4eZdZ  d dedededdffdd ZZS) NonDynamicallyQuantizableLinearTNr$r%r'rcstj|||||ddS)N)r'r*r+r)rr$r%r'r*r+rrrr~s  z(NonDynamicallyQuantizableLinear.__init__r;)rr r!r>r@rr#rrrrrB}s rBc seZdZUdZgdZeed<eed<eed<eed<  ddededed ed df fd d Z dd dZ deded efddZ d e fddZ ZS)r aApplies a bilinear transformation to the incoming data: :math:`y = x_1^T A x_2 + b` Args: in1_features: size of each first input sample in2_features: size of each second input sample out_features: size of each output sample bias: If set to False, the layer will not learn an additive bias. Default: ``True`` Shape: - Input1: :math:`(*, H_{in1})` where :math:`H_{in1}=\text{in1\_features}` and :math:`*` means any number of additional dimensions including none. All but the last dimension of the inputs should be the same. - Input2: :math:`(*, H_{in2})` where :math:`H_{in2}=\text{in2\_features}`. - Output: :math:`(*, H_{out})` where :math:`H_{out}=\text{out\_features}` and all but the last dimension are the same shape as the input. Attributes: weight: the learnable weights of the module of shape :math:`(\text{out\_features}, \text{in1\_features}, \text{in2\_features})`. The values are initialized from :math:`\mathcal{U}(-\sqrt{k}, \sqrt{k})`, where :math:`k = \frac{1}{\text{in1\_features}}` bias: the learnable bias of the module of shape :math:`(\text{out\_features})`. If :attr:`bias` is ``True``, the values are initialized from :math:`\mathcal{U}(-\sqrt{k}, \sqrt{k})`, where :math:`k = \frac{1}{\text{in1\_features}}` Examples:: >>> m = nn.Bilinear(20, 30, 40) >>> input1 = torch.randn(128, 20) >>> input2 = torch.randn(128, 30) >>> output = m(input1, input2) >>> print(output.size()) torch.Size([128, 40]) ) in1_features in2_featuresr%rCrDr%r&TNr'rcsz||d}t||_||_||_ttj|||ffi||_|r1ttj|fi||_ n| dd| dSr() rrrCrDr%rr,r-r&r'r.r/)rrCrDr%r'r*r+r0rrrrs    zBilinear.__init__cCsLdt|jd}t|j| ||jdur$t|j| |dSdS)Nr )r3r4r&sizerr5r')rr7rrrr/s  zBilinear.reset_parametersinput1input2cCst|||j|jSr)r8Zbilinearr&r')rrFrGrrrrszBilinear.forwardcCsd|j|j|j|jduS)Nz:in1_features={}, in2_features={}, out_features={}, bias={})formatrCrDr%r'r9rrrr:szBilinear.extra_reprr;r<r=rrrrr s % r csbeZdZUdZeZeed<eed<  ddede ddffdd Z dfd d Z dd d Z Z S)raA :class:`torch.nn.Linear` module where `in_features` is inferred. In this module, the `weight` and `bias` are of :class:`torch.nn.UninitializedParameter` class. They will be initialized after the first call to ``forward`` is done and the module will become a regular :class:`torch.nn.Linear` module. The ``in_features`` argument of the :class:`Linear` is inferred from the ``input.shape[-1]``. Check the :class:`torch.nn.modules.lazy.LazyModuleMixin` for further documentation on lazy modules and their limitations. Args: out_features: size of each output sample bias: If set to ``False``, the layer will not learn an additive bias. Default: ``True`` Attributes: weight: the learnable weights of the module of shape :math:`(\text{out\_features}, \text{in\_features})`. The values are initialized from :math:`\mathcal{U}(-\sqrt{k}, \sqrt{k})`, where :math:`k = \frac{1}{\text{in\_features}}` bias: the learnable bias of the module of shape :math:`(\text{out\_features})`. If :attr:`bias` is ``True``, the values are initialized from :math:`\mathcal{U}(-\sqrt{k}, \sqrt{k})` where :math:`k = \frac{1}{\text{in\_features}}` r&r'TNr%rcsL||d}tdddtdi||_||_|r$tdi||_dSdS)Nr)rFr)rrrr&r%r')rr%r'r*r+r0rrrrs zLazyLinear.__init__cs(|s|jdkrtdSdSdS)Nr)has_uninitialized_paramsr$rr/r9rrrr/szLazyLinear.reset_parameterscCs||rr@rr/rLr#rrrrrs   r)r3typingrr,rZtorch.nn.parameterrrrr8rmoduler Zlazyr __all__r rrBr rrrrrs      MJ