o hh^@sdZddlZddlZddlmZddlZddlmZdFddZdFddZ dFd d Z d d Z d dZ dFddZ   dGdedededeejdef ddZ   dGdedededeejdef ddZ     dHdedededededeejdefd d!Zded"edefd#d$Zdedefd%d&Zdedefd'd(Zd)d*ZdId,d-Zd.d/Z  dJded0edeejdefd1d2Z  dJded0edeejdefd3d4Zd5d6Z  7 8 dKdeded9ed:edeejf d;d<Z  7 8 dKdeded9ed:edeejf d=d>Z + dLdeejfd?d@Z A dMdeejfdBdCZ dDdEZ!e!eZ"e!eZ#e!eZ$e!eZ%e!eZ&e!eZ'e!eZ(e!eZ)e!eZ*e!eZ+e!e Z,dS)NzHThis file contains utilities for initializing neural network parameters.N)Optional)TensorcC<t|j|||dWdS1swYdSN generator)torchno_graduniform_tensorabrra/var/www/html/construction_image-detection-poc/venv/lib/python3.10/site-packages/torch/nn/init.py_no_grad_uniform_ $rcCrr)rr normal_r meanstdrrrr_no_grad_normal_rrc Csdd}||d|ks||d|krtjdddtD||||}||||}|jd|dd|d|d|||td| ||j ||d |WdS1sfwYdS) NcSsdt|tddS)N?@)matherfsqrt)xrrrnorm_cdfsz(_no_grad_trunc_normal_..norm_cdfzjmean is more than 2 std from [a, b] in nn.init.trunc_normal_. The distribution of values may be incorrect. stacklevelrr)minmax) warningswarnrr r erfinv_mul_rradd_clamp_) r rrr rrrlurrr_no_grad_trunc_normal_s     $r-cCs6t ||WdS1swYdSN)rr fill_r valrrr_no_grad_fill_>s $r2cCs4t |WdS1swYdSr.)rr zero_r rrr_no_grad_zero_Cs $r5cCsgd}||vs |dkrdS|dkrdS|dkrtdS|dkrM|d ur(d }nt|ts2t|ts7t|tr:|}ntd |d tdd|d S|dkrT dStd|)aReturn the recommended gain value for the given nonlinearity function. The values are as follows: ================= ==================================================== nonlinearity gain ================= ==================================================== Linear / Identity :math:`1` Conv{1,2,3}D :math:`1` Sigmoid :math:`1` Tanh :math:`\frac{5}{3}` ReLU :math:`\sqrt{2}` Leaky Relu :math:`\sqrt{\frac{2}{1 + \text{negative\_slope}^2}}` SELU :math:`\frac{3}{4}` ================= ==================================================== .. warning:: In order to implement `Self-Normalizing Neural Networks`_ , you should use ``nonlinearity='linear'`` instead of ``nonlinearity='selu'``. This gives the initial weights a variance of ``1 / N``, which is necessary to induce a stable fixed point in the forward pass. In contrast, the default gain for ``SELU`` sacrifices the normalization effect for more stable gradient flow in rectangular layers. Args: nonlinearity: the non-linear function (`nn.functional` name) param: optional parameter for the non-linear function Examples: >>> gain = nn.init.calculate_gain('leaky_relu', 0.2) # leaky_relu with negative_slope=0.2 .. _Self-Normalizing Neural Networks: https://papers.nips.cc/paper/2017/hash/5d44ee6f2c3f71b73125876103c8f6c4-Abstract.html )linearconv1dconv2dconv3dconv_transpose1dconv_transpose2dconv_transpose3dsigmoidr"tanhg?relur leaky_reluN{Gz?znegative_slope z not a valid numberrselug?zUnsupported nonlinearity )rr isinstanceboolintfloat ValueError) nonlinearityparam linear_fnsnegative_sloperrrcalculate_gainHs."  rLrr r rrreturncC4tj|rtjjt|f||||dSt||||S)aFill the input Tensor with values drawn from the uniform distribution. :math:`\mathcal{U}(a, b)`. Args: tensor: an n-dimensional `torch.Tensor` a: the lower bound of the uniform distribution b: the upper bound of the uniform distribution generator: the torch Generator to sample from (default: None) Examples: >>> w = torch.empty(3, 5) >>> nn.init.uniform_(w) r )r overrideshas_torch_function_variadichandle_torch_functionr rr rrrr r rrcCrO)aFill the input Tensor with values drawn from the normal distribution. :math:`\mathcal{N}(\text{mean}, \text{std}^2)`. Args: tensor: an n-dimensional `torch.Tensor` mean: the mean of the normal distribution std: the standard deviation of the normal distribution generator: the torch Generator to sample from (default: None) Examples: >>> w = torch.empty(3, 5) >>> nn.init.normal_(w) r)rrPrQrRrrrrrrrrSrrcCst||||||dS)aFill the input Tensor with values drawn from a truncated normal distribution. The values are effectively drawn from the normal distribution :math:`\mathcal{N}(\text{mean}, \text{std}^2)` with values outside :math:`[a, b]` redrawn until they are within the bounds. The method used for generating the random values works best when :math:`a \leq \text{mean} \leq b`. Args: tensor: an n-dimensional `torch.Tensor` mean: the mean of the normal distribution std: the standard deviation of the normal distribution a: the minimum cutoff value b: the maximum cutoff value generator: the torch Generator to sample from (default: None) Examples: >>> w = torch.empty(3, 5) >>> nn.init.trunc_normal_(w) r)r-)r rrr rrrrr trunc_normal_srUr1cCs,tj|rtjjt|f||dSt||S)zFill the input Tensor with the value :math:`\text{val}`. Args: tensor: an n-dimensional `torch.Tensor` val: the value to fill the tensor with Examples: >>> w = torch.empty(3, 5) >>> nn.init.constant_(w, 0.3) r0)rrPrQrR constant_r2r0rrrrVs   rVcCs t|dS)zFill the input Tensor with the scalar value `1`. Args: tensor: an n-dimensional `torch.Tensor` Examples: >>> w = torch.empty(3, 5) >>> nn.init.ones_(w) r)r2r4rrrones_s rWcCst|S)zFill the input Tensor with the scalar value `0`. Args: tensor: an n-dimensional `torch.Tensor` Examples: >>> w = torch.empty(3, 5) >>> nn.init.zeros_(w) )r5r4rrrzeros_s rXcCsX|dkr tdttj|j||jdWd|S1s%wY|S)a=Fill the 2-dimensional input `Tensor` with the identity matrix. Preserves the identity of the inputs in `Linear` layers, where as many inputs are preserved as possible. Args: tensor: a 2-dimensional `torch.Tensor` Examples: >>> w = torch.empty(3, 5) >>> nn.init.eye_(w) r,Only tensors with 2 dimensions are supported)out requires_gradN) ndimensionrGrr eyeshaper[r4rrreye_s   r_r"c Cs<|}|dvr td|}|d|dkrtd|d|}t||d}tg|t|D]U}t|D]N}|dkrSd||||||ddf<q<|dkrnd||||||dd|ddf<q1, each group of channels preserves identity Args: tensor: a {3, 4, 5}-dimensional `torch.Tensor` groups (int, optional): number of groups in the conv layer (default: 1) Examples: >>> w = torch.empty(3, 16, 5, 5) >>> nn.init.dirac_(w) >>> w = torch.empty(3, 24, 5, 5) >>> nn.init.dirac_(w, 3) )z5Only tensors with 3, 4, or 5 dimensions are supportedrz!dim 0 must be divisible by groupsr"r`rraN)r\rGsizer#rr r3range)r groups dimensionssizesout_chans_per_grpmin_dimgdrrrdirac_$sL    "         rlcCsp|}|dkr td|d}|d}d}|dkr,|jddD]}||9}q%||}||}||fS)NrzNFan in and fan out can not be computed for tensor with fewer than 2 dimensionsr"r)dimrGrcr^)r rfnum_input_fmapsnum_output_fmapsreceptive_field_sizesfan_infan_outrrr_calculate_fan_in_and_fan_outYs    rtgaincCsDt|\}}|tdt||}td|}t|| ||S)aFill the input `Tensor` with values using a Xavier uniform distribution. The method is described in `Understanding the difficulty of training deep feedforward neural networks` - Glorot, X. & Bengio, Y. (2010). The resulting tensor will have values sampled from :math:`\mathcal{U}(-a, a)` where .. math:: a = \text{gain} \times \sqrt{\frac{6}{\text{fan\_in} + \text{fan\_out}}} Also known as Glorot initialization. Args: tensor: an n-dimensional `torch.Tensor` gain: an optional scaling factor generator: the torch Generator to sample from (default: None) Examples: >>> w = torch.empty(3, 5) >>> nn.init.xavier_uniform_(w, gain=nn.init.calculate_gain('relu')) Note: Be aware that ``fan_in`` and ``fan_out`` are calculated assuming that the weight matrix is used in a transposed manner, (i.e., ``x @ w.T`` in ``Linear`` layers, where ``w.shape = [fan_out, fan_in]``). This is important for correct initialization. If you plan to use ``x @ w``, where ``w.shape = [fan_in, fan_out]``, pass in a transposed weight matrix, i.e. ``nn.init.xavier_uniform_(w.T, ...)``. r@)rtrrrFr)r rurrrrsrr rrrxavier_uniform_ns "rwcCs4t|\}}|tdt||}t|d||S)aFill the input `Tensor` with values using a Xavier normal distribution. The method is described in `Understanding the difficulty of training deep feedforward neural networks` - Glorot, X. & Bengio, Y. (2010). The resulting tensor will have values sampled from :math:`\mathcal{N}(0, \text{std}^2)` where .. math:: \text{std} = \text{gain} \times \sqrt{\frac{2}{\text{fan\_in} + \text{fan\_out}}} Also known as Glorot initialization. Args: tensor: an n-dimensional `torch.Tensor` gain: an optional scaling factor generator: the torch Generator to sample from (default: None) Examples: >>> w = torch.empty(3, 5) >>> nn.init.xavier_normal_(w) Note: Be aware that ``fan_in`` and ``fan_out`` are calculated assuming that the weight matrix is used in a transposed manner, (i.e., ``x @ w.T`` in ``Linear`` layers, where ``w.shape = [fan_out, fan_in]``). This is important for correct initialization. If you plan to use ``x @ w``, where ``w.shape = [fan_in, fan_out]``, pass in a transposed weight matrix, i.e. ``nn.init.xavier_normal_(w.T, ...)``. rrM)rtrrrFr)r rurrrrsrrrrxavier_normal_s !rxcCsH|}ddg}||vrtd|d|t|\}}|dkr"|S|S)NrrrszMode z" not supported, please use one of )lowerrGrt)r mode valid_modesrrrsrrr_calculate_correct_fans  r|rrr@rzrHc Cstj|rtjjt|f|||||dSd|jvr td|St||}t ||}|t |}t d|}t |j | ||dWdS1sPwYdS)aFill the input `Tensor` with values using a Kaiming uniform distribution. The method is described in `Delving deep into rectifiers: Surpassing human-level performance on ImageNet classification` - He, K. et al. (2015). The resulting tensor will have values sampled from :math:`\mathcal{U}(-\text{bound}, \text{bound})` where .. math:: \text{bound} = \text{gain} \times \sqrt{\frac{3}{\text{fan\_mode}}} Also known as He initialization. Args: tensor: an n-dimensional `torch.Tensor` a: the negative slope of the rectifier used after this layer (only used with ``'leaky_relu'``) mode: either ``'fan_in'`` (default) or ``'fan_out'``. Choosing ``'fan_in'`` preserves the magnitude of the variance of the weights in the forward pass. Choosing ``'fan_out'`` preserves the magnitudes in the backwards pass. nonlinearity: the non-linear function (`nn.functional` name), recommended to use only with ``'relu'`` or ``'leaky_relu'`` (default). generator: the torch Generator to sample from (default: None) Examples: >>> w = torch.empty(3, 5) >>> nn.init.kaiming_uniform_(w, mode='fan_in', nonlinearity='relu') Note: Be aware that ``fan_in`` and ``fan_out`` are calculated assuming that the weight matrix is used in a transposed manner, (i.e., ``x @ w.T`` in ``Linear`` layers, where ``w.shape = [fan_out, fan_in]``). This is important for correct initialization. If you plan to use ``x @ w``, where ``w.shape = [fan_in, fan_out]``, pass in a transposed weight matrix, i.e. ``nn.init.kaiming_uniform_(w.T, ...)``. )r r rzrHrr,Initializing zero-element tensors is a no-oprvrN)rrPrQrRkaiming_uniform_r^r%r&r|rLrrr r ) r r rzrHrfanrurboundrrrr~s( +    $r~cCsvd|jvr td|St||}t||}|t|}t|j d||dWdS1s4wYdS)aFill the input `Tensor` with values using a Kaiming normal distribution. The method is described in `Delving deep into rectifiers: Surpassing human-level performance on ImageNet classification` - He, K. et al. (2015). The resulting tensor will have values sampled from :math:`\mathcal{N}(0, \text{std}^2)` where .. math:: \text{std} = \frac{\text{gain}}{\sqrt{\text{fan\_mode}}} Also known as He initialization. Args: tensor: an n-dimensional `torch.Tensor` a: the negative slope of the rectifier used after this layer (only used with ``'leaky_relu'``) mode: either ``'fan_in'`` (default) or ``'fan_out'``. Choosing ``'fan_in'`` preserves the magnitude of the variance of the weights in the forward pass. Choosing ``'fan_out'`` preserves the magnitudes in the backwards pass. nonlinearity: the non-linear function (`nn.functional` name), recommended to use only with ``'relu'`` or ``'leaky_relu'`` (default). generator: the torch Generator to sample from (default: None) Examples: >>> w = torch.empty(3, 5) >>> nn.init.kaiming_normal_(w, mode='fan_out', nonlinearity='relu') Note: Be aware that ``fan_in`` and ``fan_out`` are calculated assuming that the weight matrix is used in a transposed manner, (i.e., ``x @ w.T`` in ``Linear`` layers, where ``w.shape = [fan_out, fan_in]``). This is important for correct initialization. If you plan to use ``x @ w``, where ``w.shape = [fan_in, fan_out]``, pass in a transposed weight matrix, i.e. ``nn.init.kaiming_normal_(w.T, ...)``. rr}rN) r^r%r&r|rLrrrr r)r r rzrHrrrurrrrkaiming_normal_ s +    $rc Cs|dkr td|dkr|S|d}||}|||fjdd|d}||kr2|tj |\}}t |d}| } || 9}||krP|t | ||||Wd|S1smwY|S)aFill the input `Tensor` with a (semi) orthogonal matrix. Described in `Exact solutions to the nonlinear dynamics of learning in deep linear neural networks` - Saxe, A. et al. (2013). The input tensor must have at least 2 dimensions, and for tensors with more than 2 dimensions the trailing dimensions are flattened. Args: tensor: an n-dimensional `torch.Tensor`, where :math:`n \geq 2` gain: optional scaling factor generator: the torch Generator to sample from (default: None) Examples: >>> # xdoctest: +REQUIRES(env:TORCH_DOCTEST_LAPACK) >>> w = torch.empty(3, 5) >>> nn.init.orthogonal_(w) rz4Only tensors with 2 or more dimensions are supportedrr"rN)r\rGnumelrc new_emptyrt_rlinalgqrdiagsignr view_ascopy_r() r rurrowscols flattenedqrrkphrrr orthogonal_>s,        rrAc Cs|dkr td|j\}}tt||}t)|jd||dt |D]}t |}|d|} d|| |f<q)Wd|S1sHwY|S)aFill the 2D input `Tensor` as a sparse matrix. The non-zero elements will be drawn from the normal distribution :math:`\mathcal{N}(0, 0.01)`, as described in `Deep learning via Hessian-free optimization` - Martens, J. (2010). Args: tensor: an n-dimensional `torch.Tensor` sparsity: The fraction of elements in each column to be set to zero std: the standard deviation of the normal distribution used to generate the non-zero values generator: the torch Generator to sample from (default: None) Examples: >>> w = torch.empty(3, 5) >>> nn.init.sparse_(w, sparsity=0.1) rrYrrN) r\rGr^rErceilrr rrdrandperm) r sparsityrrrr num_zeroscol_idx row_indices zero_indicesrrrsparse_qs       rcsFjddfdd}dddd|_|_|S)Ncs,tjdddtdd|i|S)Nz `nn.init.z)` is now deprecated in favor of `nn.init.z`.rr )r%r& FutureWarning)argskwargsmethnew_nameold_namerrdeprecated_inits z(_make_deprecate..deprecated_initz z_(...) .. warning:: This method is now deprecated in favor of :func:`torch.nn.init.z"`. See :func:`~torch.nn.init.z` for details.)__name____doc__)rrrrr_make_deprecates  rr.)rMrN)rMrrTrN)r")rN)rrrr@N)r"N)rAN)-rrr%typingr _Optionalrrrrr-r2r5rLrF Generatorr rrUrVrWrXr_rlrtrwrxr|strr~rrrruniformnormalconstantr]diracxavier_uniform xavier_normalkaiming_uniformkaiming_normal orthogonalsparserrrrs     % H     5 + '  C 7 6 '