+ jD^RIt^RIHt^RIHtHt^RIt^RIHt^RIH t ^RI H t ^RI H t ^RIHtHtHt^RIHt^R IHtHt.R/OtR0t!RR]P24tRRltR1RRlltRRlt!RR ]PP24tRR2R]RR/t!RR ]4t !RR ]4t!!RR ]4t"!RR]4t#R R!lt$]!4]!R"] PJ3R#7R$RR%R&/R'R(ll44t&]!4]!R"]!PJ3R#7R$RR%R&/R)R*ll44t']!4]!R"]"PJ3R#7R$RR%R&/R+R,ll44t(]!4]!R"]#PJ3R#7R$RR%R&/R-R.ll44t)R#)3N)partial)AnyOptionalTensor)ImageClassification)_log_api_usage_once)register_modelWeights WeightsEnum)_IMAGENET_CATEGORIES)_ovewrite_named_paramhandle_legacy_interfaceMNASNetMNASNet0_5_WeightsMNASNet0_75_WeightsMNASNet1_0_WeightsMNASNet1_3_WeightscTaa]tRt^"toRV3RlV3RllltV3RlRltRtVtV;t#)_InvertedResidualcB<V^8dQhRS[RS[RS[RS[RS[RS[RR/#) in_chout_ch kernel_sizestrideexpansion_factor bn_momentumreturnN)intfloat)format __classdict__s"[/home/luca/screen-checking/.venv/lib/python3.14/site-packages/torchvision/models/mnasnet.py __annotate___InvertedResidual.__annotate__#sE   "% 47 AD X[ jo  cd<\SV`4VR 9d\RV 24hVR 9d\RV 24hW,pW8H;'dV^8HVn\P !\P !W^RR7\P!WvR7\P!RR7\P !WwW3^,WGRR7\P!WvR7\P!RR7\P !Wr^RR7\P!W&R74Vn R #) z#stride should be 1 or 2 instead of z(kernel_size should be 3 or 5 instead of F)biasmomentumTinplacepaddingrgroupsr*Nr)r)) super__init__ ValueErrorapply_residualnn SequentialConv2d BatchNorm2dReLUlayers) selfrrrrrrmid_ch __class__s &&&&&&& r$r6_InvertedResidual.__init__#s   B6(KL L f $G }UV V)#o==&A+mm IIeQU 3 NN6 8 GGD ! IIfk!;KTZpu v NN6 8 GGD ! IIfae 4 NN6 8   r'c&<V^8dQhRS[RS[/#)rinputrr)r"r#s"r$r%r&;s&&V&&r'cxVP'dVPV4V,#VPV4#)Nr8r>)r?rDs&&r$forward_InvertedResidual.forward;s1    ;;u%- -;;u% %r'rF)g?) __name__ __module__ __qualname____firstlineno__r6rG__static_attributes____classdictcell__ __classcell__rAr#s@@r$rr"s  0&&&r'rcV^8dQhR\R\R\R\R\R\R\R\P/#) rrrrr exp_factorrepeatsrr)r r!r9r:)r"s"r$r%r%BsP , ,  , ,*- ,7: ,HK ,VY ,hm ,]] ,r'c V^8d\RV 24h\WW#WFR7p.p\^V4F!p VP\WV^WFR74K# \P !V.VO5!#)z&Creates a stack of inverted residuals.z$repeats should be >= 1, instead got )r)r7rrangeappendr9r:) rrrrrRrSrfirst remaining_s &&&&&&& r$_stackrZBsp{?yIJJ e[* fEI 1g *6;:op == + ++r'cHV^8dQhR\R\R\R\/#)rvaldivisor round_up_biasrr!r )r"s"r$r%r%Ps/LLuLsL5LSVLr'cRTu;8dR8gM\RV 24h\V\W^, ,4V,V,4pW2V,8dV#W1,#)zAsymmetric rounding to make `val` divisible by `divisor`. With default bias, will round up, unless the number is no more than 10% greater than the smaller divisible value, i.e. (83, 8) -> 80, but (84, 8) -> 88.?zIround_up_bias should be greater than 0.0 and smaller than 1.0 instead of )r7maxr )r\r]r^new_vals&&& r$_round_to_multiple_ofrePs`  $ $derdstuu'3sq[01W>> model = MNASNet(1.0, num_classes=1000) >>> x = torch.rand(1, 3, 224, 224) >>> y = model(x) >>> y.dim() 2 >>> y.nelement() 1000 c0<V^8dQhRS[RS[RS[RR/#)rrg num_classesdropoutrNr_)r"r#s"r$r%MNASNet.__annotate__ps*,','e,'#,'u,'W[,'r'c<\SV`4\V4VR8:d\RV 24hWnW n\ V4p\P!^V^,^^^RR7\P!V^,\R7\P!RR7\P!V^,V^,^^^V^,RR7\P!V^,\R7\P!RR7\P!V^,V^,^^^RR7\P!V^,\R7\V^,V^,^^^^\4\V^,V^,^^^^\4\V^,V^,^^^^\4\V^,V^,^^^^\4\V^,V^,^^^^\4\V^,V^,^^^^\4\P!V^,R^^^RR7\P!R\R7\P!RR7.p\P!V!Vn\P!\P!VRR 7\P !RV44VnVP%4EFxp\'V\P4'dh\P(P+VP,R R R 7VP.e,\P(P1VP.4KK\'V\P4'dV\P(P3VP,4\P(P1VP.4EK\'V\P 4'gEK#\P(P5VP,R RR 7\P(P1VP.4EK{ R #)raz,alpha should be greater than 0.0 instead of Fr0rr*r+Tr-r/i)pr.fan_outrelu)mode nonlinearityNsigmoid)r5r6r r7rgrvrsr9r;r< _BN_MOMENTUMr=rZr:r>DropoutLinear classifiermodules isinstanceinitkaiming_normal_weightr*zeros_ones_kaiming_uniform_)r?rgrvrwrqr>mrAs&&&& r$r6MNASNet.__init__ps D! C<KE7ST T &U# IIaAq G NN6!9| < GGD ! IIfQiAq6RS9[` a NN6!9| < GGD ! IIfQiAq O NN6!9| < 6!9fQiAq!\ B 6!9fQiAq!\ B 6!9fQiAq!\ B 6!9fQiAq!\ B 6!9fQiAq!\ B 6!9fQiAq!\ B IIfQiq!AE J NN4, 7 GGD !+ .mmV, -- Wd(KRYYW[]hMijA!RYY''''yv'V66%GGNN166*&Ar~~.. ahh'qvv&Aryy))(( PY(Zqvv& r'c&<V^8dQhRS[RS[/#)rxrr)r"r#s"r$r%rxs"""F"r'clVPV4pVP^^.4pVPV4#)r)r>meanr)r?rs&&r$rGMNASNet.forwards/ KKN FFAq6Nq!!r'cx<V^8dQhRS[RS[RS[RS[RS[S[,RS[S[,RS[S[,RR /#) r state_dictprefixlocal_metadatastrict missing_keysunexpected_keys error_msgsrN)dictstrboolrh)r"r#s"r$r%rxsg/ / / /  /  / 3i / c/ I/  / r'c8<VPRR4pVR 9d\RV 24hV^8XEdXVPR8XEgF\VP4p \P !^^ ^^^RR7\P !^ \R7\P!RR7\P !^ ^ ^^^^ RR 7\P !^ \R7\P!RR7\P !^ ^^^^RR7\P !^\R7\^V ^,^^^^\4. p \V 4FwrWPV &K ^Vn \P!R \4\ S V`EWW4WVV4R#) versionNz+version should be set to 1 or 2 instead of rbFrzr+Tr-r/aA new version of MNASNet model has been implemented. Your checkpoint was saved using the previous version. This checkpoint will load and work as before, but you may want to upgrade by training a newer model or transfer learning from an updated ImageNet checkpoint.r2)getr7rgrsr9r;r<rr=rZ enumerater>_versionwarningswarn UserWarningr5_load_from_state_dict)r?rrrrrrrrrqv1_stemidxlayerrAs&&&&&&&& r$rMNASNet._load_from_state_dictsQ!$$Y5 & J7)TU U a< c 1 !,F !RAaeDrL9% "b!Qq%PrL9% "b!QquErL9r6!9aAq,? G(0 #( C 1DM MMI    %  Wa r')rrgrr>rv)ig?) rIrJrKrL__doc__rr6rGrrMrNrOrPs@@r$rras5 H,','\"" / / / r'min_size categoriesrecipez(https://github.com/1e100/mnasnet_trainerch]tRt^t]!R]!]^R7/]CRRRRRRR R //R R R RRR/CR7t]t Rt R#)rzIhttps://download.pytorch.org/models/mnasnet0.5_top1_67.823-3ffadce67e.pth crop_size num_paramsi!_metrics ImageNet-1Kacc@1g"P@acc@5g(\U@_opsg9v? _file_sizeg;O.!@_docs9These weights reproduce closely the results of the paper.url transformsmetaN rIrJrKrLr rr _COMMON_META IMAGENET1K_V1DEFAULTrMrr'r$rrsd W.#>    ' VV  E % T  M"Gr'cn]tRt^t]!R]!]^^R7/]CRRRRRRR R R R //R RRRRR/CR7t]t Rt R#)rz`_. rrNrrr'r$rrsn J.#3O  C ' VV  E &  M*Gr'ch]tRtRt]!R]!]^R7/]CRRRRRR R R //R R RRRR/CR7t]t Rt R#)ri zIhttps://download.pytorch.org/models/mnasnet1.0_top1_73.512-f206786ef8.pthrriPBrrrgw/]R@rgq= ףV@rgjt?rg ףp=0@rrrrNrrr'r$rr sd W.#>    ' VV  E & T  M"Gr'cn]tRtRt]!R]!]^^R7/]CRRRRRR R R R R //RRRRRR/CR7t]t Rt R#)ri z;https://download.pytorch.org/models/mnasnet1_3-a4c69d6f.pthrrrri_rrrgMb S@rg rhaW@rgE?rg">8@rrrrNrrr'r$rr sn I.#3O  C ' VV  E &  M*Gr'c jV^8dQhR\R\\,R\R\R\ /#)rrgweightsprogresskwargsr)r!rr rrr)r"s"r$r%r%9s6  E H[$9 T UX ]d r'c Ve(\VR\VPR,44\V3/VBpV'd#VP VP VRR74V#)NrvrT)r check_hash)rlenrrload_state_dictget_state_dict)rgrrrmodels&&&, r$_mnasnetr9sXfmSl9S5TU E $V $E g44hSW4XY Lr' pretrained)rrrTc^V^8dQhR\\,R\R\R\/#rrrrr)rrrrr)r"s"r$r%r%G0668$676$6ad6ip6r'c F\PV4p\RW3/VB#)anMNASNet with depth multiplier of 0.5 from `MnasNet: Platform-Aware Neural Architecture Search for Mobile `_ paper. Args: weights (:class:`~torchvision.models.MNASNet0_5_Weights`, optional): The pretrained weights to use. See :class:`~torchvision.models.MNASNet0_5_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.mnasnet.MNASNet`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.MNASNet0_5_Weights :members: g?)rverifyrrrrs$$,r$ mnasnet0_5rE&.!''0G C 5f 55r'c^V^8dQhR\\,R\R\R\/#r)rrrrr)r"s"r$r%r%cs077H%897D7cf7kr7r'c F\PV4p\RW3/VB#)arMNASNet with depth multiplier of 0.75 from `MnasNet: Platform-Aware Neural Architecture Search for Mobile `_ paper. Args: weights (:class:`~torchvision.models.MNASNet0_75_Weights`, optional): The pretrained weights to use. See :class:`~torchvision.models.MNASNet0_75_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.mnasnet.MNASNet`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.MNASNet0_75_Weights :members: g?)rrrrs$$,r$ mnasnet0_75ras&."((1G D' 6v 66r'c^V^8dQhR\\,R\R\R\/#r)rrrrr)r"s"r$r%r%rr'c F\PV4p\RW3/VB#)anMNASNet with depth multiplier of 1.0 from `MnasNet: Platform-Aware Neural Architecture Search for Mobile `_ paper. Args: weights (:class:`~torchvision.models.MNASNet1_0_Weights`, optional): The pretrained weights to use. See :class:`~torchvision.models.MNASNet1_0_Weights` below for more details, and possible values. By default, no pre-trained weights are used. progress (bool, optional): If True, displays a progress bar of the download to stderr. Default is True. **kwargs: parameters passed to the ``torchvision.models.mnasnet.MNASNet`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.MNASNet1_0_Weights :members: rb)rrrrs$$,r$ mnasnet1_0r}rr'c^V^8dQhR\\,R\R\R\/#r)rrrrr)r"s"r$r%r%rr'c F\PV4p\RW3/VB#)anMNASNet with depth multiplier of 1.3 from `MnasNet: Platform-Aware Neural Architecture Search for Mobile `_ paper. Args: weights (:class:`~torchvision.models.MNASNet1_3_Weights`, optional): The pretrained weights to use. See :class:`~torchvision.models.MNASNet1_3_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.mnasnet.MNASNet`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.MNASNet1_3_Weights :members: g?)rrrrs$$,r$ mnasnet1_3rrr') rrrrrrrrrg0U0*3?)g?)r)r))*r functoolsrtypingrrtorchtorch.nnr9rtransforms._presetsrutilsr _apir r r _metar _utilsrr__all__rModulerrZrersrrrrrrrrrrrrrr'r$rs 5'66'B  & &@ ,LIr ehhoor l& 8 *+2*2 ,0B0P0P!QR6$6QU6S64,0C0Q0Q!RS7D7SW7T74,0B0P0P!QR6$6QU6S64,0B0P0P!QR6$6QU6S6r'