+ jnB ^RIHt^RIHtHtHt^RIt^RIHt^RIH t ^RI H t ^RI H t ^RIHtHtHt^RIHt^R IHtHt^R I HtHtHtHt^R IHtHtHt.R:Ot!RR] P@4t!!RR ] PD4t#RRlt$RR;R]RRRRRR/t%!RR ]4t&!RR]4t'!R R]4t(!R!R]4t)]!R"R#7]!R$R%3R&7R'RR(R)R*R+/R,R-ll44t*]!R.R#7]!R$R/3R&7R'RR(R)R*R+/R0R1ll44t+]!R2R#7]!R$R33R&7R'RR(R)R*R+/R4R5ll44t,]!R6R#7]!R$R73R&7R'RR(R)R*R+/R8R9ll44t-R#)<)partial)AnyOptionalUnionNTensor) shufflenetv2)ImageClassification)register_modelWeights WeightsEnum)_IMAGENET_CATEGORIES)_ovewrite_named_paramhandle_legacy_interface)ShuffleNet_V2_X0_5_WeightsShuffleNet_V2_X1_0_WeightsShuffleNet_V2_X1_5_WeightsShuffleNet_V2_X2_0_Weights) _fuse_modules _replace_reluquantize_modelQuantizableShuffleNetV2#ShuffleNet_V2_X0_5_QuantizedWeights#ShuffleNet_V2_X1_0_QuantizedWeights#ShuffleNet_V2_X1_5_QuantizedWeights#ShuffleNet_V2_X2_0_QuantizedWeightscPaa]tRt^#toV3RlV3RlltV3RlRltRtVtV;t#)QuantizableInvertedResidualc*<V^8dQhRS[RS[RR/#argskwargsreturnNr)format __classdict__s"m/home/luca/screen-checking/.venv/lib/python3.14/site-packages/torchvision/models/quantization/shufflenetv2.py __annotate__(QuantizableInvertedResidual.__annotate__$s"22c2S2T2cn<\SV`!V/VB\PP 4VnR#N)super__init__nn quantizedFloatFunctionalcatselfr"r# __class__s&*,r(r/$QuantizableInvertedResidual.__init__$s) $)&)<<//1r+c&<V^8dQhRS[RS[/#r!xr$r)r&r's"r(r)r*(s   F r+cVVP^8XdDVP^^R7wr#VPPW PV4.^R7pM=VPPVP V4VPV4.^R7p\ P !V^4pV#))dim)stridechunkr3branch2branch1r channel_shuffle)r5r:x1x2outs&& r(forward#QuantizableInvertedResidual.forward(s ;;! WWQAW&FB((,,LL$451,=C((,, QaAq,IC**32 r+)r3) __name__ __module__ __qualname____firstlineno__r/rF__static_attributes____classdictcell__ __classcell__r6r's@@r(rr#s22   r+rcfaa]tRt^4toV3RlV3RlltV3RlRltRV3RlRlltRtVtV;t #) rc*<V^8dQhRS[RS[RR/#r r%)r&r's"r(r)$QuantizableShuffleNetV2.__annotate__6s";;c;S;T;r+c<\SV`!VR\/VB\PP P 4Vn\PP P4Vn R#)inverted_residualN) r.r/rtorchao quantization QuantStubquant DeQuantStubdequantr4s&*,r(r/ QuantizableShuffleNetV2.__init__6sP $X2MXQWXXX**446 xx,,88: r+c&<V^8dQhRS[RS[/#r9r)r&r's"r(r)rR;sFr+clVPV4pVPV4pVPV4pV#r-)rY _forward_implr[)r5r:s&&r(rFQuantizableShuffleNetV2.forward;s1 JJqM   q ! LLOr+c4<V^8dQhRS[S[,RR/#)r!is_qatr$N)rbool)r&r's"r(r)rRAs$4r+c VPP4F&wr#VR9gKVfK\V.R O.VRR7K( VP4Fp\ V4\ JgK\ VPPP44^8d \VPRR..R O.VRR7\VP.R ORR..R O.VRR7K R#) aFuse conv/bn/relu modules in shufflenetv2 model Fuse conv+bn+relu/ conv+relu/conv+bn modules to prepare for quantization. Model is modified in place. .. note:: Note that this operation does not change numerics and the model after modification is in floating point N01T)inplace34)conv1conv5)rerf2)rlrhri)567) _modulesitemsrmodulestyperlenrAr@)r5rbnamems&& r( fuse_model"QuantizableShuffleNetV2.fuse_modelAs}}**,GD))ama/!2FDI-AAw55qyy))//12Q6!!))sCj/-JF\`aII$sCj/B   r+)r[rYr-) rHrIrJrKr/rFrwrLrMrNrOs@@r(rr4s(;;  r+cV^8dQhR\\,R\\,R\\,R\R\R\ R\ /#)r!stages_repeatsstages_out_channelsweightsprogressquantizer#r$)listintrr rcrr)r&s"r(r)r)Zs[Ick "     r+c ~VeW\VR\VPR,44RVP9d\VRVPR,4VPRR4p\ W3/VBp\ V4V'd \ Wv4Ve#VPVPVRR74V#)N num_classes categoriesbackendfbgemmT)r} check_hash) rrtmetapoprrrload_state_dictget_state_dict)rzr{r|r}r~r#rmodels&&$$$, r( _shufflenetv2rZsfmSl9S5TU  $ !&)W\\)5L MjjH-G #N R6 RE%u& g44hSW4XY Lr+min_sizerrrrecipezdhttps://github.com/pytorch/vision/tree/main/references/classification#post-training-quantized-models_docsz These weights were produced by doing Post Training Quantization (eager mode) on top of the unquantized weights listed below. c|]tRt^t]!R]!]^R7/]CRRR]PRRRR R R //R R RR/CR7t ] t Rt R#)rzShttps://download.pytorch.org/models/quantized/shufflenetv2_x0.5_fbgemm-00845098.pth crop_size num_paramsi unquantized_metrics ImageNet-1Kacc@1g#~jL@acc@5gRS@_opsg{Gz? _file_sizegjt?url transformsrN) rHrIrJrKr rr _COMMON_METAr IMAGENET1K_V1IMAGENET1K_FBGEMM_V1DEFAULTrLrr+r(rrsk" a.#>    ' 5CC VV  D %  "#Gr+c|]tRt^t]!R]!]^R7/]CRRR]PRRRR R R //R R RR/CR7t ] t Rt R#)rzQhttps://download.pytorch.org/models/quantized/shufflenetv2_x1_fbgemm-1e62bb32.pthrri"rrrrgףp= Q@rgh|?U@rg(\?rgy&1@rrN) rHrIrJrKr rr rrrrrrLrr+r(rrsk" _.#>    ' 5CC VV  E %  "#Gr+c]tRt^t]!R]!]^^R7/]CRRRRR]PRR R R R R //RRRR/CR7t ] t Rt R#)rzShttps://download.pytorch.org/models/quantized/shufflenetv2_x1_5_fbgemm-d7401f05.pthr resize_sizer+https://github.com/pytorch/vision/pull/5906riv5rrrrgSR@rg̬V@rgl?rgK7A` @rrN) rHrIrJrKr rr rrrrrrLrr+r(rrs" a.#3O    C ' 5CC VV  E %  $#Gr+c]tRt^t]!R]!]^^R7/]CRRRRR]PRR R R R R //RRRR/CR7t ] t Rt R#)rzShttps://download.pytorch.org/models/quantized/shufflenetv2_x2_0_fbgemm-5cac526c.pthrrrriprrrrg-R@rgZd;W@rg-?rg|?5@rrN) rHrIrJrKr rr rrrrrrLrr+r(rrrr+quantized_shufflenet_v2_x0_5)ru pretrainedcrVPRR4'd\P#\P#r~F)getrrrrr#s&r(r1zz*e,, 0 D D :+99 :r+)r|r|r}Tr~Fc V^8dQhR\\\\3,,R\R\R\ R\ /#r!r|r}r~r#r$)rrrrrcrr)r&s"r(r)r)M** e?A[[\ ]** *  *  *r+c vV'd\M\PV4p\.RO.RO3RVRVRV/VB#)a Constructs a ShuffleNetV2 with 0.5x output channels, as described in `ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design `__. .. note:: Note that ``quantize = True`` returns a quantized model with 8 bit weights. Quantized models only support inference and run on CPUs. GPU inference is not yet supported. Args: weights (:class:`~torchvision.models.quantization.ShuffleNet_V2_X0_5_QuantizedWeights` or :class:`~torchvision.models.ShuffleNet_V2_X0_5_Weights`, optional): The pretrained weights for the model. See :class:`~torchvision.models.quantization.ShuffleNet_V2_X0_5_QuantizedWeights` 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. quantize (bool, optional): If True, return a quantized version of the model. Default is False. **kwargs: parameters passed to the ``torchvision.models.quantization.ShuffleNet_V2_X0_5_QuantizedWeights`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.quantization.ShuffleNet_V2_X0_5_QuantizedWeights :members: .. autoclass:: torchvision.models.ShuffleNet_V2_X0_5_Weights :members: :noindex: r|r}r~r)0`)rrverifyrr|r}r~r#s$$$,r(shufflenet_v2_x0_5rsOd7?2D^ffgnoG * 4; FN Ya ek r+quantized_shufflenet_v2_x1_0crVPRR4'd\P#\P#r)rrrrrrs&r(rrrr+c V^8dQhR\\\\3,,R\R\R\ R\ /#r)rrrrrcrr)r&s"r(r)r)rr+c vV'd\M\PV4p\.RO.RO3RVRVRV/VB#)a Constructs a ShuffleNetV2 with 1.0x output channels, as described in `ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design `__. .. note:: Note that ``quantize = True`` returns a quantized model with 8 bit weights. Quantized models only support inference and run on CPUs. GPU inference is not yet supported. Args: weights (:class:`~torchvision.models.quantization.ShuffleNet_V2_X1_0_QuantizedWeights` or :class:`~torchvision.models.ShuffleNet_V2_X1_0_Weights`, optional): The pretrained weights for the model. See :class:`~torchvision.models.quantization.ShuffleNet_V2_X1_0_QuantizedWeights` 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. quantize (bool, optional): If True, return a quantized version of the model. Default is False. **kwargs: parameters passed to the ``torchvision.models.quantization.ShuffleNet_V2_X1_0_QuantizedWeights`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.quantization.ShuffleNet_V2_X1_0_QuantizedWeights :members: .. autoclass:: torchvision.models.ShuffleNet_V2_X1_0_Weights :members: :noindex: r|r}r~r)rtir)rrrrrs$$$,r(shufflenet_v2_x1_0rOd7?2D^ffgnoG , 6= HP [c gm r+quantized_shufflenet_v2_x1_5crVPRR4'd\P#\P#r)rrrrrrs&r(rrJrr+c V^8dQhR\\\\3,,R\R\R\ R\ /#r)rrrrrcrr)r&s"r(r)r)Qrr+c vV'd\M\PV4p\.RO.RO3RVRVRV/VB#)a Constructs a ShuffleNetV2 with 1.5x output channels, as described in `ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design `__. .. note:: Note that ``quantize = True`` returns a quantized model with 8 bit weights. Quantized models only support inference and run on CPUs. GPU inference is not yet supported. Args: weights (:class:`~torchvision.models.quantization.ShuffleNet_V2_X1_5_QuantizedWeights` or :class:`~torchvision.models.ShuffleNet_V2_X1_5_Weights`, optional): The pretrained weights for the model. See :class:`~torchvision.models.quantization.ShuffleNet_V2_X1_5_QuantizedWeights` below for more details, and possible values. By default, no pre-trained weights are used. progress (bool, optional): If True, displays a progress bar of the download to stderr. Default is True. quantize (bool, optional): If True, return a quantized version of the model. Default is False. **kwargs: parameters passed to the ``torchvision.models.quantization.ShuffleNet_V2_X1_5_QuantizedWeights`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.quantization.ShuffleNet_V2_X1_5_QuantizedWeights :members: .. autoclass:: torchvision.models.ShuffleNet_V2_X1_5_Weights :members: :noindex: r|r}r~r)ri`ir)rrrrrs$$$,r(shufflenet_v2_x1_5rFrr+quantized_shufflenet_v2_x2_0crVPRR4'd\P#\P#r)rrrrrrs&r(rrrr+c V^8dQhR\\\\3,,R\R\R\ R\ /#r)rrrrrcrr)r&s"r(r)r)rr+c vV'd\M\PV4p\.RO.RO3RVRVRV/VB#)a Constructs a ShuffleNetV2 with 2.0x output channels, as described in `ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design `__. .. note:: Note that ``quantize = True`` returns a quantized model with 8 bit weights. Quantized models only support inference and run on CPUs. GPU inference is not yet supported. Args: weights (:class:`~torchvision.models.quantization.ShuffleNet_V2_X2_0_QuantizedWeights` or :class:`~torchvision.models.ShuffleNet_V2_X2_0_Weights`, optional): The pretrained weights for the model. See :class:`~torchvision.models.quantization.ShuffleNet_V2_X2_0_QuantizedWeights` 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. quantize (bool, optional): If True, return a quantized version of the model. Default is False. **kwargs: parameters passed to the ``torchvision.models.quantization.ShuffleNet_V2_X2_0_QuantizedWeights`` base class. Please refer to the `source code `_ for more details about this class. .. autoclass:: torchvision.models.quantization.ShuffleNet_V2_X2_0_QuantizedWeights :members: .. autoclass:: torchvision.models.ShuffleNet_V2_X2_0_Weights :members: :noindex: r|r}r~r)riii)rrrrrs$$$,r(shufflenet_v2_x2_0r~rr+) rrrrrrrrr)r<r<). functoolsrtypingrrrrUtorch.nnr0rtorchvision.modelsr transforms._presetsr _apir r r _metar_utilsrrrrrrutilsrrr__all__InvertedResidualr ShuffleNetV2rrrrrrrrrrrrr+r(rs'' +677(C @? ,"?"?"#l77#L6& x t   #+#*#+#*#+#,#+#,34   *`d** * 5*Z34   *`d** * 5*Z34   *`d** * 5*Z34   *`d** * 5*r+