3jSSKJr SSKJrJrJr SSKJr SSKJ r SSK J r "SS\R\ \ 5r "S S \R\ \ 5rg ) ) annotations)AnyLiteralOptional)nn)ImageModuleMixIn)ONNXExportMixinc^\rSrSrSrS U4Sjjr\S Sj5r\RS SSjj5rSSS.SU4S jjjr S r U=r $) ImageModuleavHandles image-based operations. This modules accepts multiple input and output data types, provides end-to-end visualization, file saving features. Note that this module fits the classes that return one image tensor only. Note: The additional add-on features increase the use of memories. To restore the original behaviour, you may set `disable_features = True`. c4>[TU]"U0UD6 SUlgNFsuper__init___disable_featuresselfargskwargs __class__s L/home/wildlama/miniconda3/lib/python3.13/site-packages/kornia/core/module.pyrImageModule.__init__) $)&)',cUR$)aWhether convenience I/O (input/output) helper features are disabled. This flag controls the extra behavior provided by :class:`ImageModuleMixIn`, such as automatic input conversion (for example, PIL or NumPy inputs), automatic output conversion, and output caching for visualization helpers. Returns: ``True`` if the helper features are bypassed and the module behaves like a plain :class:`torch.nn.Module` call. ``False`` if helper features are enabled. rrs rdisable_featuresImageModule.disable_features-%%%rcXlg)a)Enable or disable convenience I/O (input/output) handling features. Args: value: Feature toggle. - ``True``: disable automatic type conversion and output caching, so ``__call__`` behaves closer to a raw PyTorch module call. - ``False``: keep helper features active. Nrrvalues rr r!; "'rNptinput_names_to_handle output_typec>UR(dLURXS9"[TU] 5nU"U0UD6nUS:XaUR U5UlU$X`lU$[TU] "U0UD6nU$aoOverwrite the __call__ function to handle various inputs. Args: inputs: Inputs to operate on. input_names_to_handle: List of input names to convert, if None, handle all inputs. output_type: Desired output type ('pt', 'numpy', or 'pil'). kwargs: Additional arguments. Returns: Callable: Decorated function with converted input and output types. r(r'rconvert_input_outputr__call___detach_tensor_to_cpu _output_imagerr)r*inputsrdecorated_forwardr1rs rr/ImageModule.__call__G(%% $ 9 9&;!:!g! .v@@Md"%)%?%? %N" &3""G,f??Mrrr1rrrrreturnNoner9boolTr%r<r9r: r3rr)zOptional[list[Any]]r*zLiteral['pt', 'numpy', 'pil']rrr9r __name__ __module__ __qualname____firstlineno____doc__rpropertyr setterr/__static_attributes__ __classcell__rs@rr r } - & & ' '6:59  33    rr c^\rSrSrSrS U4Sjjr\S Sj5r\RS SSjj5rSSS.SU4S jjjr S r U=r $)ImageSequentialiaHandles image-based operations as a sequential module. This modules accepts multiple input and output data types, provides end-to-end visualization, file saving features. Note that this module fits the classes that return one image tensor only. Note: The additional add-on features increase the use of memories. To restore the original behaviour, you may set `disable_features = True`. c4>[TU]"U0UD6 SUlgrrrs rrImageSequential.__init__vrrcUR$)aWhether convenience I/O (input/output) helper features are disabled. This flag controls the extra behavior provided by :class:`ImageModuleMixIn`, such as automatic input conversion (for example, PIL or NumPy inputs), automatic output conversion, and output caching for visualization helpers. Returns: ``True`` if the helper features are bypassed and the module behaves like a plain :class:`torch.nn.Sequential` call. ``False`` if helper features are enabled. rrs rr ImageSequential.disable_featureszr"rcXlg)a-Enable or disable convenience I/O (input/output) handling features. Args: value: Feature toggle. - ``True``: disable automatic type conversion and output caching, so ``__call__`` behaves closer to a raw PyTorch sequential call. - ``False``: keep helper features active. Nrr$s rr rRr&rNr'r(c>UR(dLURXS9"[TU] 5nU"U0UD6nUS:XaUR U5UlU$X`lU$[TU] "U0UD6nU$r,r-r2s rr/ImageSequential.__call__r6rr7r8r;r=r>r?r@rJs@rrMrMirKrrMN) __future__rtypingrrrtorchrmixin.image_moduler mixin.onnxr Moduler SequentialrMrrr^sH$#))0'J"))-JZJbmm%5Jr