j( UddlmZddlmZddlmZddlmZddlm Z ddl m Z m Z gdZ d d d eed d dd d d eed ddd d d eed dddZeeeeeffed<GddejZ ddededeedzdedef dZd dededefdZd dededefdZd dededefdZdS)!)deepcopy)Any)nn)VOCABS FASTConvLayer)conv_sequence_ptload_pretrained_params) textnet_tiny textnet_small textnet_base)gh|?5?g= ףp=?gV-?)gA`"?gl?g$C?r rfrenchzOhttps://doctr-static.mindee.com/models?id=v0.8.1/textnet_tiny-27288d12.pt&src=0)meanstd input_shapeclassesurlzPhttps://doctr-static.mindee.com/models?id=v0.8.1/textnet_small-43166ee6.pt&src=0zOhttps://doctr-static.mindee.com/models?id=v0.8.1/textnet_base-7f68d7e0.pt&src=0 default_cfgsc eZdZdZ ddeeeeefdeeeefded e d eee fdzd df fd Z d ede d dfdZ xZ S)TextNetaImplements TextNet architecture from `"FAST: Faster Arbitrarily-Shaped Text Detector with Minimalist Kernel Representation" `_. Implementation based on the official Pytorch implementation: `_. Args: stages (list[dict[str, list[int]]]): list of dictionaries containing the parameters of each stage. include_top (bool, optional): Whether to include the classifier head. Defaults to True. num_classes (int, optional): Number of output classes. Defaults to 1000. cfg (dict[str, Any], optional): Additional configuration. Defaults to None. rTNstagesr num_classes include_topcfgreturnc gtdddddddd|D}|rr|tjtjdtjdtj|d d d |tj|||_ | D]}t|tj r(tj |jd d Dt|tjrJtj |jdtj |jddS)Nr@T)r#) in_channels out_channelsrelubn kernel_sizestridepaddingc g|]FtjdfdttdDDGS)c&g|]}tdi|S)r).0paramss n/var/www/html/Carbon-Document/venv/lib/python3.11/site-packages/doctr/models/classification/textnet/pytorch.py z/TextNet.__init__...Es6   "++F++   c0g|]fdDS)c.i|]}||Sr-r-)r.keyistages r0 z:TextNet.__init__....Gs##H#H#H3CsA#H#H#Hr2r-)r.r6r7s @r0r1z/TextNet.__init__...Gs2"s"s"sQ#H#H#H#H#H%#H#H#H"s"s"sr2r$)r Sequentialrangelen)r.r7s @r0r1z$TextNet.__init__..Ds{     "s"s"s"sRWX[\abo\pXqXqRrRr"s"s"s   r2r#r%fan_outr&)mode nonlinearityr)r appendrr9AdaptiveAvgPool2dFlattenLinearsuper__init__rmodules isinstanceConv2dinitkaiming_normal_weight BatchNorm2d constant_bias) selfrrrrr_layersm __class__s r0rEzTextNet.__init__8sx $ BTdPQZ[ek $  $  $    NN (++JqMMIfRj8`_. Implementation based on the official Pytorch implementation: `_. >>> import torch >>> from doctr.models import textnet_tiny >>> model = textnet_tiny(pretrained=False) >>> input_tensor = torch.rand((1, 3, 512, 512), dtype=torch.float32) >>> out = model(input_tensor) Args: pretrained: boolean, True if model is pretrained **kwargs: keyword arguments of the TextNet architecture Returns: A textnet tiny model r r!rrr)r#r"r#r$r%r(r))r!rororo)rmr#rrmrr#)r"r#r#r#)rorsrsrs)rmrmrrrq)rsrtrtrt)rmrrrqrm 7.2.weight7.2.biasrrdrkrcrTs r0r r s$   D1HrdQhPVxZ[|gpgpgp q q222!$ ???&,,    433!$ ???&,,    433!$ ???&,,    *":.1  2 3  r2c td|fdgdzdgdzdgdzddgdgddgd zgd gd dgd d gd zgdgd dgddgdzgdgddgddgd|S)aImplements TextNet architecture from `"FAST: Faster Arbitrarily-Shaped Text Detector with Minimalist Kernel Representation" `_. Implementation based on the official Pytorch implementation: `_. >>> import torch >>> from doctr.models import textnet_small >>> model = textnet_small(pretrained=False) >>> input_tensor = torch.rand((1, 3, 512, 512), dtype=torch.float32) >>> out = model(input_tensor) Args: pretrained: boolean, True if model is pretrained **kwargs: keyword arguments of the TextNet architecture Returns: A TextNet small model r r!r"rmr#rn)r!rorororororororo)rmrqrmrrrmrrrqrmr"r#r#r#r#r#r#r#rorsrsrsrsrsrsrsrs)rmrmrqrrrmrqrrrmrsrtrtrtrtrt)rmrrrqrqrrr"r#r#r#r#rurvrwrxrys r0r r s$   D1HrdQhPVxZ[|hiklgm n nFFF!$ ___222    HGG!$ ___222    988!$ GGG)//    *":.1  2 3  r2c td|fdgdzdgdzgdgddgddgdzgd gd dgd d gd zgdgddgddgdzgdgddgddgd|S)aImplements TextNet architecture from `"FAST: Faster Arbitrarily-Shaped Text Detector with Minimalist Kernel Representation" `_. Implementation based on the official Pytorch implementation: `_. >>> import torch >>> from doctr.models import textnet_base >>> model = textnet_base(pretrained=False) >>> input_tensor = torch.rand((1, 3, 512, 512), dtype=torch.float32) >>> out = model(input_tensor) Args: pretrained: boolean, True if model is pretrained **kwargs: keyword arguments of the TextNet architecture Returns: A TextNet base model rr! ) rmrmrrrmrrrmrmrqrmrm) r#r"r#r#r#r#r#r#r#r#rn) r!rorororororororororo) rmrqrmrrrmrmrrrrrmrm) r"r#r#r#r#r#r#r#r#r#r}rsr{)rmrmrmrqrmrrrmrrr|r~rtr)rmrqrrrrrqrrurvrwrxrys r0rrs$   !#tby!#r ooo888    QPP!$ ooo888    HGG!$ ___222    988!$ GGG)//   ' 4":.;  < =  r2)N)F)copyrtypingrtorchrdoctr.datasetsrmodules.layersr utilsr r __all__r[rr\r]__annotations__r9rr`rkr r rr-r2r0rsL!!!!!!++++++======== ; ; ;&$"4x())` &$"4x())a &$"4x())` ++ d3S#X&'2;<;<;<;<;