o h_O@sddlZddlZddlmZddlZddlmZddZddZdEdd Z d d Z d d Z dEddZ dFdede de defddZdFdede de defddZ     dGdede de de de deejdefdd Zded!e defd"d#Zdedefd$d%Zdedefd&d'Zd(d)ZdHd+d,Zd-d.ZdIded/e defd0d1ZdIded/e defd2d3Zd4d5Z 7dJdede d8ed9efd:d;Z 7dJdede d8ed9efdd?ZdKdAdBZdCdDZ e eZ!e eZ"e eZ#e eZ$e eZ%e eZ&e eZ'e eZ(e eZ)e eZ*e eZ+dS)LN)Tensor)OptionalcC8t|||WdS1swYdSN)torchno_graduniform_tensorabr C/var/www/html/ai/venv/lib/python3.10/site-packages/torch/nn/init.py_no_grad_uniform_   $rcCrr)rrnormal_r meanstdr r r_no_grad_normal_rrc Csdd}||d|ks||d|krtjdddtD||||}||||}|jd|dd|d|d|||td| ||j ||d |WdS1sfwYdS) NcSsdt|tddS)N?@)matherfsqrt)xr r rnorm_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. stacklevel generatorr)minmax) warningswarnrrrerfinv_mul_rradd_clamp_) r rrr r r"rlur r r_no_grad_trunc_normal_s    $r-cCs6t ||WdS1swYdSr)rrfill_r valr r r_no_grad_fill_9s $r1cCs4t |WdS1swYdSr)rrzero_r r r r_no_grad_zero_>s $r4cCsgd}||vs |dkrdS|dkrdS|dkrtdS|dkrM|d ur(d }nt|ts2t|ts7t|tr:|}ntd |d tdd|d S|dkrSdStd|)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_sloper r rcalculate_gainCs"! rKrr r r returncC0tj|rtjjt|f|||dSt|||S)adFills 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 Examples: >>> w = torch.empty(3, 5) >>> nn.init.uniform_(w) r )r overrideshas_torch_function_variadichandle_torch_functionrrr r r rrz  rrrcCrN)azFills 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 Examples: >>> w = torch.empty(3, 5) >>> nn.init.normal_(w) r)rrOrPrQrrrr r rrrRrrr"cCst||||||dS)aFills 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 Examples: >>> w = torch.empty(3, 5) >>> nn.init.trunc_normal_(w) r!)r-)r rrr r r"r r r trunc_normal_srTr0cCs,tj|rtjjt|f||dSt||S)zFills 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) r/)rrOrPrQ constant_r1r/r r rrUs  rUcCs t|dS)zFills 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)r1r3r r rones_s rVcCst|S)zFills 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) )r4r3r r rzeros_s rWcCsX|dkr tdttj|j||jdWd|S1s%wY|S)a=Fills 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) ndimensionrFrreyeshaperZr3r r reye_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_rr`N)r[rFsizer#rrr2range)r groups dimensionssizesout_chans_per_grpmin_dimgdr r rdirac_s:    "    rkcCsp|}|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)dimrFrbr])r renum_input_fmapsnum_output_fmapsreceptive_field_sizesfan_infan_outr r r_calculate_fan_in_and_fan_out#s    rsgaincCsBt|\}}|tdt||}td|}t|| |S)aFills the input `Tensor` with values according to the method described in `Understanding the difficulty of training deep feedforward neural networks` - Glorot, X. & Bengio, Y. (2010), using a uniform distribution. 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 Examples: >>> w = torch.empty(3, 5) >>> nn.init.xavier_uniform_(w, gain=nn.init.calculate_gain('relu')) r@)rsrrrEr)r rtrqrrrr r r rxavier_uniform_6s rvcCs2t|\}}|tdt||}t|d|S)aFills the input `Tensor` with values according to the method described in `Understanding the difficulty of training deep feedforward neural networks` - Glorot, X. & Bengio, Y. (2010), using a normal distribution. 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 Examples: >>> w = torch.empty(3, 5) >>> nn.init.xavier_normal_(w) rrL)rsrrrEr)r rtrqrrrr r rxavier_normal_Qs  rwcCsH|}ddg}||vrtd|d|t|\}}|dkr"|S|S)NrqrrzMode z" not supported, please use one of )lowerrFrs)r mode valid_modesrqrrr r r_calculate_correct_fanks  r{rqr?ryrGcCstj|rtjjt|f||||dSd|jvrtd|St||}t ||}|t |}t d|}t | | |WdS1sMwYdS)aFills the input `Tensor` with values according to the method described in `Delving deep into rectifiers: Surpassing human-level performance on ImageNet classification` - He, K. et al. (2015), using a uniform distribution. 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). Examples: >>> w = torch.empty(3, 5) >>> nn.init.kaiming_uniform_(w, mode='fan_in', nonlinearity='relu') )r r ryrGr,Initializing zero-element tensors is a no-opruN)rrOrPrQkaiming_uniform_r]r%r&r{rKrrrr)r r ryrGfanrtrboundr r rr}us&       $r}cCsrd|jvr td|St||}t||}|t|}t| d|WdS1s2wYdS)aFills the input `Tensor` with values according to the method described in `Delving deep into rectifiers: Surpassing human-level performance on ImageNet classification` - He, K. et al. (2015), using a normal distribution. 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). Examples: >>> w = torch.empty(3, 5) >>> nn.init.kaiming_normal_(w, mode='fan_out', nonlinearity='relu') rr|N) r]r%r&r{rKrrrrr)r r ryrGr~rtrr r rkaiming_normal_s      $rc Cs|dkr td|dkr|S|d}||}|||dd}||kr/|tj |\}}t |d}| }||9}||krM|t | ||||Wd|S1sjwY|S)a]Fills the input `Tensor` with a (semi) orthogonal matrix, as 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 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 N)r[rFnumelrbnewrt_rlinalgqrdiagsignrview_ascopy_r() r rtrowscols flattenedqrrjphr r r orthogonal_s,        rr@c Cs|dkr td|j\}}tt||}t'|d|t |D]}t |}|d|}d|||f<q'Wd|S1sFwY|S)aNFills the 2D input `Tensor` as a sparse matrix, where 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 Examples: >>> w = torch.empty(3, 5) >>> nn.init.sparse_(w, sparsity=0.1) rrXrN) r[rFr]rDrceilrrrrcrandperm) r sparsityrrr num_zeroscol_idx row_indices zero_indicesr r rsparse_s        rcsFjddfdd}dddd|_|_|S)Ncs*tjddddd|i|S)Nznn.init.z' is now deprecated in favor of nn.init..rr)r%r&)argskwargsmethnew_nameold_namer rdeprecated_initsz(_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__)rrr rr_make_deprecates  rr)rLr)rLrrSrN)r )r)rrqr?)r@),rr%rrtypingr _Optionalrrr-r1r4rKrErr GeneratorrTrUrVrWr^rkrsrvrwr{strr}rrrruniformnormalconstantr\diracxavier_uniform xavier_normalkaiming_uniformkaiming_normal orthogonalsparser r r rs   # 7   +  2 ' -