YhA>ddlZddlmZddlmZddlmZmZddlZddl m Z ddl m cm Z ddlmcmZddlmZddlmZddlmZd d lmZmZmZd d lmZd d lmZmZgd ZGdde j Z!Gdde j"Z#Gdde j$Z%Gdde j Z&de j dede'ddfdZ(de)de*e)e)e)e)fde)deede'dede&fdZ+d ed!d"d#Z,Gd$d%eZ-Gd&d'eZ.Gd(d)eZ/Gd*d+eZ0eed,e-j1f-dd.d/dee-de'dede&fd0Z2eed,e.j1f-dd.d/dee.de'dede&fd1Z3eed,e/j1f-dd.d/dee/de'dede&fd2Z4eed,e0j1f-dd.d/dee0de'dede&fd3Z5dS)4N) OrderedDict)partial)AnyOptional)Tensor)ImageClassification)_log_api_usage_once)register_modelWeights WeightsEnum)_IMAGENET_CATEGORIES)_ovewrite_named_paramhandle_legacy_interface) DenseNetDenseNet121_WeightsDenseNet161_WeightsDenseNet169_WeightsDenseNet201_Weights densenet121 densenet161 densenet169 densenet201c 8eZdZ ddedededededdf fd Zd eedefd Z d eedefd Z e j j d eedefdZe j jd eedefdZe j jd edefdZd edefdZxZS) _DenseLayerFnum_input_features growth_ratebn_size drop_ratememory_efficientreturnNcttj||_tjd|_tj|||zddd|_tj||z|_ tjd|_ tj||z|dddd|_ t||_ ||_dS)NTinplacer F kernel_sizestridebiasr'r(paddingr))super__init__nn BatchNorm2dnorm1ReLUrelu1Conv2dconv1norm2relu2conv2floatr r!)selfrrrr r! __class__s m/var/www/tools.fuzzalab.pt/emblema-extractor/venv/lib/python3.11/site-packages/torchvision/models/densenet.pyr.z_DenseLayer.__init__ s ^$677 WT*** Y17[3HVW`ahmnnn ^Gk$9:: WT*** Yw4kqYZdelqrrr y)) 0inputsctj|d}||||}|SNr )torchcatr5r3r1)r:r>concated_featuresbottleneck_outputs r< bn_functionz_DenseLayer.bn_function/sE!Ifa00 JJtzz$**=N2O2O'P'PQQ  r=inputc$|D] }|jrdS dS)NTF) requires_grad)r:rFtensors r<any_requires_gradz_DenseLayer.any_requires_grad5s-  F# tt ur=c8fd}tj|g|RddiS)Nc.|SN)rE)r>r:s r<closurez7_DenseLayer.call_checkpoint_bottleneck..closure=s##F++ +r= use_reentrantF)cp checkpoint)r:rFrNs` r<call_checkpoint_bottleneckz&_DenseLayer.call_checkpoint_bottleneck;s> , , , , ,}WBuBBBEBBBr=cdSrMr:rFs r<forwardz_DenseLayer.forwardB r=cdSrMrTrUs r<rVz_DenseLayer.forwardFrWr=ct|tr|g}n|}|jrX||rCtjrtd||}n| |}| | | |}|j dkr!tj||j |j}|S)Nz%Memory Efficient not supported in JITr)ptraining) isinstancerr!rJrAjit is_scripting ExceptionrRrEr8r7r6r Fdropoutr[)r:rF prev_featuresrD new_featuress r<rVz_DenseLayer.forwardLs eV $ $ ""GMM!M   @T%;%;M%J%J @y%%'' I GHHH $ ? ? N N   $ 0 0 ? ? zz$**TZZ8I-J-J"K"KLL >A  9\T^dm\\\Lr=F)__name__ __module__ __qualname__intr9boolr.listrrErJrAr]unusedrR_overload_methodrV __classcell__r;s@r<rrsrw 1 1"% 147 1BE 1RW 1ko 1  1 1 1 1 1 1!$v,!6!!!! tF|  YCV CCCCC  Y T&\ f      Y V      Vr=rcTeZdZdZ ddedededededed d ffd Zd ed efd Z xZ S) _DenseBlockrF num_layersrrrr r!r"Nctt|D]8}t|||zz||||}|d|dzz|9dS)N)rrr r!z denselayer%dr )r-r.ranger add_module) r:rqrrrr r!ilayerr;s r<r.z_DenseBlock.__init__cs z"" = =A"Q_4'#!1 E OONa!e4e < < < < = =r= init_featuresc|g}|D]%\}}||}||&tj|dSr@)itemsappendrArB)r:rwfeaturesnamervrcs r<rVz_DenseBlock.forwardwsX!?::<< * *KD% 5??L OOL ) ) ) )y1%%%r=rd) rerfrg_versionrhr9rir.rrVrmrns@r<rprp`sH"'=== = =  =  == ======(&V&&&&&&&&&r=rpc,eZdZdededdffd ZxZS) _Transitionrnum_output_featuresr"Nc ttj||_tjd|_tj||ddd|_tj dd|_ dS)NTr$r Fr&r)r'r() r-r.r/r0normr2relur4conv AvgPool2dpool)r:rrr;s r<r.z_Transition.__init__sz N#566 GD))) I02EST]^ejkkk LQq999 r=)rerfrgrhr.rmrns@r<rrsR:3:S:T::::::::::r=rcxeZdZdZ dd ed eeeeefd ed ed edededdffd Zde de fdZ xZ S)raKDensenet-BC model class, based on `"Densely Connected Convolutional Networks" `_. Args: growth_rate (int) - how many filters to add each layer (`k` in paper) block_config (list of 4 ints) - how many layers in each pooling block num_init_features (int) - the number of filters to learn in the first convolution layer bn_size (int) - multiplicative factor for number of bottle neck layers (i.e. bn_size * k features in the bottleneck layer) drop_rate (float) - dropout rate after each dense layer num_classes (int) - number of classification classes memory_efficient (bool) - If True, uses checkpointing. Much more memory efficient, but slower. Default: *False*. See `"paper" `_.  @rFr block_confignum_init_featuresrr num_classesr!r"Nc :tt|tjt dtjd|ddddfdtj|fdtjd fd tj ddd fg|_ |}t|D]\} } t| |||||} |j d| d zz| || |zz}| t|d z kr:t||dz} |j d| d zz| |dz}|j dtj|tj|||_|D]} t'| tjr%tj| jAt'| tjrKtj| jd tj| jdt'| tjr%tj| jddS)Nconv0r*rFr+norm0relu0Tr$pool0r )r'r(r,)rqrrrr r!z denseblock%d)rrz transition%dnorm5r)r-r.r r/ Sequentialrr4r0r2 MaxPool2dr{ enumeraterprtlenrLinear classifiermodulesr\initkaiming_normal_weight constant_r))r:rrrrr rr! num_featuresrurqblocktransmr;s r<r.zDenseNet.__init__s D!!! bi+??@bgd3334blqANNNO       ) &|44 1 1MAz%#/'#!1 E M $ $^q1u%=u E E E'*{*BBLC %%)))#|YeijYjkkk ((1q5)A5III+q0    ".*F*FGGG)L+>> - -A!RY'' -''1111Ar~.. -!!!(A...!!!&!,,,,Ary)) -!!!&!,,, - -r=xc||}tj|d}tj|d}t j|d}||}|S)NTr$)r r r )r{r`radaptive_avg_pool2drAflattenr)r:rr{outs r<rVzDenseNet.forwards`==##fXt,,,#C00mC##ooc"" r=)rrrrrrF) rerfrg__doc__rhtupler9rir.rrVrmrns@r<rrs  "2A!#!&:-:-:-Cc3./:- :-  :-  :-:-:- :-:-:-:-:-:-xFr=rmodelweightsprogressr"cntjd}||d}t|D]R}||}|r9|d|dz}||||<||=S||dS)Nz]^(.*denselayer\d+\.(?:norm|relu|conv))\.((?:[12])\.(?:weight|bias|running_mean|running_var))$T)r check_hashr r)recompileget_state_dictrjkeysmatchgroupload_state_dict)rrrpattern state_dictkeyresnew_keys r<_load_state_dictrs jhG''d'KKJJOO%%&&  mmC    iillSYYq\\1G",S/Jw 3 *%%%%%r=rrrkwargsc |)t|dt|jdt|||fi|}|t ||||S)Nr categories)rrr)rrmetarr)rrrrrrrs r< _densenetrsefmSl9S5T5TUUU [,0A L LV L LEugIIII Lr=)rz*https://github.com/pytorch/vision/pull/116z'These weights are ported from LuaTorch.)min_sizerrecipe_docsc `eZdZedeediedddddid d d  ZeZd S)rz@ num_params_metrics_ops _file_sizeurl transformsrN rerfrgr rr _COMMON_META IMAGENET1K_V1DEFAULTrTr=r<rr sG J7.#>>>    !##        M GGGr=rc `eZdZedeediedddddid d d  ZeZd S)rz>>    "##   !    M GGGr=rc `eZdZedeediedddddid d d  ZeZd S)rz>>    "##        M GGGr=rc `eZdZedeediedddddid d d  ZeZd S)rz>>    "##        M GGGr=r pretrained)rT)rrc Xt|}tddd||fi|S)a{Densenet-121 model from `Densely Connected Convolutional Networks `_. Args: weights (:class:`~torchvision.models.DenseNet121_Weights`, optional): The pretrained weights to use. See :class:`~torchvision.models.DenseNet121_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.densenet.DenseNet`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.DenseNet121_Weights :members: rrr)rverifyrrrrs r<rr[5*"((11G R"gx J J6 J JJr=c Xt|}tddd||fi|S)a{Densenet-161 model from `Densely Connected Convolutional Networks `_. Args: weights (:class:`~torchvision.models.DenseNet161_Weights`, optional): The pretrained weights to use. See :class:`~torchvision.models.DenseNet161_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.densenet.DenseNet`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.DenseNet161_Weights :members: 0)rr$r`)rrrrs r<rrurr=c Xt|}tddd||fi|S)a{Densenet-169 model from `Densely Connected Convolutional Networks `_. Args: weights (:class:`~torchvision.models.DenseNet169_Weights`, optional): The pretrained weights to use. See :class:`~torchvision.models.DenseNet169_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.densenet.DenseNet`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.DenseNet169_Weights :members: r)rrrrr)rrrrs r<rrrr=c Xt|}tddd||fi|S)a{Densenet-201 model from `Densely Connected Convolutional Networks `_. Args: weights (:class:`~torchvision.models.DenseNet201_Weights`, optional): The pretrained weights to use. See :class:`~torchvision.models.DenseNet201_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.densenet.DenseNet`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.DenseNet201_Weights :members: r)rrrrr)rrrrs r<rrrr=)6r collectionsr functoolsrtypingrrrAtorch.nnr/torch.nn.functional functionalr`torch.utils.checkpointutilsrQrPrtransforms._presetsr r _apir r r_metar_utilsrr__all__Moduler ModuleDictrprrrrirrhrrrrrrrrrrrrrTr=r<rsv ###### #########555555''''''6666666666''''''BBBBBBBB   >>>>>")>>>B&&&&&"-&&&>:::::"-:::RRRRRryRRRj&BI& &t&PT&&&&&S#s*+k "     (&: :  +(+(+(+(,0C0Q!RSSS<@SWKKKH%89KDKcfKksKKKTSK0,0C0Q!RSSS<@SWKKKH%89KDKcfKksKKKTSK0,0C0Q!RSSS<@SWKKKH%89KDKcfKksKKKTSK0,0C0Q!RSSS<@SWKKKH%89KDKcfKksKKKTSKKKr=