h; SSKJr SSKJr SSKJrJrJr SSKrSSKJ r J r SSK J r SSKJrJr SS KJr SS KJr SS KJr S S KJrJrJr S SKJr S SKJrJr /SQr "SS\ RB5r""SS\ RF5r$"SS5r%"SS\ RF5r&S\'\%S\(S\\S\)S\S\&4 Sjr*S \S!S"S#.r+"S$S%\5r,"S&S'\5r-"S(S)\5r."S*S+\5r/\"5\"S,\,R`4S-9SS.S/.S\\,S\)S\S\&4S0jj55r1\"5\"S,\-R`4S-9SS.S/.S\\-S\)S\S\&4S1jj55r2\"5\"S,\.R`4S-9SS.S/.S\\.S\)S\S\&4S2jj55r3\"5\"S,\/R`4S-9SS.S/.S\\/S\)S\S\&4S3jj55r4g)4)Sequence)partial)AnyCallableOptionalN)nnTensor) functional)Conv2dNormActivationPermute)StochasticDepth)ImageClassification)_log_api_usage_once)register_modelWeights WeightsEnum)_IMAGENET_CATEGORIES)_ovewrite_named_paramhandle_legacy_interface) ConvNeXtConvNeXt_Tiny_WeightsConvNeXt_Small_WeightsConvNeXt_Base_WeightsConvNeXt_Large_Weights convnext_tinyconvnext_small convnext_baseconvnext_largec&\rSrSrS\S\4SjrSrg) LayerNorm2dxreturncURSSSS5n[R"XRURUR UR 5nURSSSS5nU$)Nrr r)permuteF layer_normnormalized_shapeweightbiasepsselfr$s S/var/www/auris/envauris/lib/python3.13/site-packages/torchvision/models/convnext.pyforwardLayerNorm2d.forward sU IIaAq ! LL114;; 488 T IIaAq !N)__name__ __module__ __qualname____firstlineno__r r2__static_attributes__r5r4r1r"r"sFr4r"c x^\rSrSrS S\S\S\\S\R4SS4U4Sjjjr S \ S\ 4S jr S r U=r $) CNBlock'N layer_scalestochastic_depth_prob norm_layer.r%c  >[TU]5 Uc[[RSS9n[R "[R "XSSUSS9[/SQ5U"U5[R"USU-SS 9[R"5[R"SU-USS 9[/S Q55Ul [R"[R"US S 5U-5Ul[US 5Ulg) Nư>r.r'T) kernel_sizepaddinggroupsr-)rr r'r) in_features out_featuresr-)rr'rr rrow)super__init__rr LayerNorm SequentialConv2dr LinearGELUblock Parametertorchonesr>rstochastic_depth)r0dimr>r?r@ __class__s r1rMCNBlock.__init__(s    48J]] IIcAq4 P L ! sO II#AG$ G GGI II!c'$ G L !  << 31(= (KL /0Eu Mr4inputclURURU5-nURU5nX!- nU$N)r>rSrW)r0r[results r1r2CNBlock.forward?s7!!DJJu$55&&v. r4)rSr>rWr])r6r7r8r9floatrrrModulerMr r2r: __classcell__rYs@r1r<r<'sj :> NN % N Xc299n56 N  NN.Vr4r<cB\rSrSrS\S\\S\SS4SjrS\4SjrS r g) CNBlockConfigFinput_channels out_channels num_layersr%Nc(XlX lX0lgr])rgrhri)r0rgrhris r1rMCNBlockConfig.__init__Hs -($r4cURRS-nUS- nUS- nUS- nUS- nUR"S0URD6$)N(zinput_channels={input_channels}z, out_channels={out_channels}z, num_layers={num_layers})r5)rYr6format__dict__)r0ss r1__repr__CNBlockConfig.__repr__RsV NN # #c ) .. ,, (( Sxx($--((r4)rgrirh) r6r7r8r9intrrMstrrrr:r5r4r1rereFs=%%sm% %  %)#)r4rec^\rSrSrSS\\S\S\S\S\\ S\ R4S \\ S\ R4S \ S S4U4S jjjr S \S \4SjrS \S \4SjrSrU=r$)r[N block_settingr?r> num_classesrS.r@kwargsr%c 4>[TU]5 [U5 U(d [S5e[ U[ 5(a/[ UVs/sHn[ U[5PM sn5(d [S5eUc[nUc[[SS9n/n USRn U R[SU SSSUSSS 95 [S U55n Sn UHn /n[!U R"5H5nX,-U S - - nURU"U RUU55 U S - n M7 U R[$R&"U65 U R(cMU R[$R&"U"U R5[$R*"U RU R(S S S955 M [$R&"U 6Ul[$R."S 5UlUSnUR(b UR(O URn[$R&"U"U5[$R2"S 5[$R4"UU55UlUR95Hn[ U[$R*[$R445(dM4[$R:R=UR>SS9 UR@cMk[$R:RCUR@5 M gs snf)Nz%The block_setting should not be emptyz/The block_setting should be List[CNBlockConfig]rBrCrr'rHT)rEstriderFr@activation_layerr-c38# UHoRv M g7fr])ri).0cnfs r1 $ConvNeXt.__init__..s I=C=sg?rr )rEr|g{Gz?)std)"rLrMr ValueError isinstancerallre TypeErrorr<rr"rgappendr sumrangerirrOrhrPfeaturesAdaptiveAvgPool2davgpoolFlattenrQ classifiermodulesinit trunc_normal_r,r-zeros_)r0rxr?r>ryrSr@rzrqlayersfirstconv_output_channelstotal_stage_blocksstage_block_idrstage_sd_prob lastblocklastconv_output_channelsmrYs r1rMConvNeXt.__init__\s D!DE E]H55#er>ser`az!]?[er>s:t:tMN N =E   $7J"$%2!$4$C$C! )%!%  ! I= II C%'E3>>*/@DVY\D\] U3#5#5{GLM!# + MM"--/ 0+ MM"3#5#56 #"4"4c6F6FTU^_`!$ v. ++A. !"% &/&<&<&HI " "iNfNf !-- / 0"**Q-KcepAq A!bii344%%ahhD%966%GGNN166*  s?ts Lr$clURU5nURU5nURU5nU$r])rrrr/s r1 _forward_implConvNeXt._forward_impls0 MM!  LLO OOA r4c$URU5$r])rr/s r1r2ConvNeXt.forwards!!!$$r4)rrr)grBiNN)r6r7r8r9listrer`rtrrrrarrMr rr2r:rbrcs@r1rr[s(+!489=L+M*L+ %L+ L+  L+ bii01 L+Xc299n56L+L+ L+L+\v& %%F%%r4rrxr?weightsprogressrzr%c Ub#[US[URS55 [U4SU0UD6nUbUR UR USS95 U$)Nry categoriesr?T)r check_hash)rlenmetarload_state_dictget_state_dict)rxr?rrrzmodels r1 _convnextrsdfmSl9S5TU ] Z:O ZSY ZE g44hSW4XY Lr4) rzNhttps://github.com/pytorch/vision/tree/main/references/classification#convnexta These weights improve upon the results of the original paper by using a modified version of TorchVision's `new training recipe `_. )min_sizerrecipe_docsc P\rSrSr\"S\"\SSS90\ESSSS S .0S S S .ES9r\r Sr g)rz>https://download.pytorch.org/models/convnext_tiny-983f1562.pth crop_size resize_sizeiH< ImageNet-1KgzGT@gMbX X@zacc@1zacc@5gm@gV-G[@ num_params_metrics_ops _file_sizeurl transformsrr5N r6r7r8r9rrr _COMMON_META IMAGENET1K_V1DEFAULTr:r5r4r1rrsS L.#3O    "##  !  M Gr4rc P\rSrSr\"S\"\SSS90\ESSSSS .0S S S .ES 9r\r Sr g)rz?https://download.pytorch.org/models/convnext_small-0c510722.pthrriHZrgClT@g)X@rg|?5^!@g"~g@rrr5Nrr5r4r1rrsS M.#3O    "##  !  M Gr4rc P\rSrSr\"S\"\SSS90\ESSSS S .0S S S .ES9r\r Sr g)rz>https://download.pytorch.org/models/convnext_base-6075fbad.pthrrihGrgU@gHz7X@rg(\µ.@g/$!u@rrr5Nrr5r4r1rrsS L.#3O    "##  !  M Gr4rc P\rSrSr\"S\"\SSS90\ESSSS S .0S S S .ES9r\r Sr g)riz?https://download.pytorch.org/models/convnext_large-ea097f82.pthrrri rg"~U@gX9v>X@rg|?5.A@gK@rrr5Nrr5r4r1rrsS M.#3O    ###  !  M Gr4r pretrained)rT)rrc [RU5n[SSS5[SSS5[SSS5[SSS5/nURSS 5n[ X4X40UD6$) aMConvNeXt Tiny model architecture from the `A ConvNet for the 2020s `_ paper. Args: weights (:class:`~torchvision.models.convnext.ConvNeXt_Tiny_Weights`, optional): The pretrained weights to use. See :class:`~torchvision.models.convnext.ConvNeXt_Tiny_Weights` below for more details and possible values. By default, no pre-trained weights are used. progress (bool, optional): If True, displays a progress bar of the download to stderr. Default is True. **kwargs: parameters passed to the ``torchvision.models.convnext.ConvNext`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.ConvNeXt_Tiny_Weights :members: `r' Nr?g?)rverifyrepoprrrrzrxr?s r1rr"st&$**73G b#q!c3"c3"c4# M #JJ'>D ]7 WPV WWr4c [RU5n[SSS5[SSS5[SSS5[SSS5/nURSS 5n[ X4X40UD6$) aQConvNeXt Small model architecture from the `A ConvNet for the 2020s `_ paper. Args: weights (:class:`~torchvision.models.convnext.ConvNeXt_Small_Weights`, optional): The pretrained weights to use. See :class:`~torchvision.models.convnext.ConvNeXt_Small_Weights` below for more details and possible values. By default, no pre-trained weights are used. progress (bool, optional): If True, displays a progress bar of the download to stderr. Default is True. **kwargs: parameters passed to the ``torchvision.models.convnext.ConvNext`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.ConvNeXt_Small_Weights :members: rrr'rrNr?g?)rrrerrrs r1rrAst*%++G4G b#q!c3"c3#c4# M #JJ'>D ]7 WPV WWr4c [RU5n[SSS5[SSS5[SSS5[SSS5/nURSS 5n[ X4X40UD6$) aMConvNeXt Base model architecture from the `A ConvNet for the 2020s `_ paper. Args: weights (:class:`~torchvision.models.convnext.ConvNeXt_Base_Weights`, optional): The pretrained weights to use. See :class:`~torchvision.models.convnext.ConvNeXt_Base_Weights` below for more details and possible values. By default, no pre-trained weights are used. progress (bool, optional): If True, displays a progress bar of the download to stderr. Default is True. **kwargs: parameters passed to the ``torchvision.models.convnext.ConvNext`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.ConvNeXt_Base_Weights :members: r'iirNr??)rrrerrrs r1rrbst&$**73G c3"c3"c4$dD!$ M #JJ'>D ]7 WPV WWr4c [RU5n[SSS5[SSS5[SSS5[SSS5/nURSS 5n[ X4X40UD6$) aQConvNeXt Large model architecture from the `A ConvNet for the 2020s `_ paper. Args: weights (:class:`~torchvision.models.convnext.ConvNeXt_Large_Weights`, optional): The pretrained weights to use. See :class:`~torchvision.models.convnext.ConvNeXt_Large_Weights` below for more details and possible values. By default, no pre-trained weights are used. progress (bool, optional): If True, displays a progress bar of the download to stderr. Default is True. **kwargs: parameters passed to the ``torchvision.models.convnext.ConvNext`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.ConvNeXt_Large_Weights :members: rrr'rirNr?r)rrrerrrs r1r r st*%++G4G c3"c3"c4$dD!$ M #JJ'>D ]7 WPV WWr4)5collections.abcr functoolsrtypingrrrrUrr torch.nnr r)ops.miscr r ops.stochastic_depthrtransforms._presetsrutilsr_apirrr_metar_utilsrr__all__rNr"rar<rerrr`boolrrrrrrrrrrr r5r4r1rs$** $425'66'B ",,bii>))*V%ryyV%r & k "     &&^  K([(K([(,0E0S0S!TU@DW[Xh'<=XPTXgjXowXVX:,0F0T0T!UV37$X/0XCGXZ]X XWX>,0E0S0S!TU@DW[Xh'<=XPTXgjXowXVX:,0F0T0T!UV37$X/0XCGXZ]X XWXr4