j UddlmZddlmZddlmZddlmZddlm Z dgZ gd Z e e ed <d ed ed ede fdZ dd ed ededed ede f dZdS))Any) PreProcessor)_CompiledModule) recognition)RecognitionPredictorrecognition_predictor) crnn_vgg16_bncrnn_mobilenet_v3_smallcrnn_mobilenet_v3_large sar_resnet31master vitstr_small vitstr_baseparseq viptr_tinyARCHSarch pretrainedkwargsreturnc ,t|trM|tvrtd|dt j|||dd}ntjtjtj tj tj tj tg}t|t|stdt||}|dd|d|jd|d<|d|jd|d<|d d |d <|jd d d}t%t'|fd di||}|S)Nzunknown architecture ''pretrained_backboneT)rrzunknown architecture: meanstd batch_size input_shapepreserve_aspect_ratio) isinstancestrr ValueErrorr__dict__getCRNNSARMASTERViTSTRPARSeqVIPTRrtupletypepopcfgr r)rrr_model allowed_archsr predictors _/var/www/html/Carbon-Document/venv/lib/python3.11/site-packages/doctr/models/recognition/zoo.py _predictorr6s$ u  =d===>> >%d+!vzzBWY]7^7^      O          $m 4 455 DBd4jjBBCC C JJ$d+++ZZ 6(:;;F6NJJufj&788F5M!::lC88F<*]+BCC0K$\+%d%dUY%d]c%d%dflmmI r Fr symmetric_padrc $td||||d|S)aText recognition architecture. Example:: >>> import numpy as np >>> from doctr.models import recognition_predictor >>> model = recognition_predictor(pretrained=True) >>> input_page = (255 * np.random.rand(32, 128, 3)).astype(np.uint8) >>> out = model([input_page]) Args: arch: name of the architecture or model itself to use (e.g. 'crnn_vgg16_bn') pretrained: If True, returns a model pre-trained on our text recognition dataset symmetric_pad: if True, pad the image symmetrically instead of padding at the bottom-right batch_size: number of samples the model processes in parallel **kwargs: optional parameters to be passed to the architecture Returns: Recognition predictor )rrr8r)r6)rrr8rrs r5r r As%4  u4Jm`j u unt u uur7N)r FFr)typingrdoctr.models.preprocessorrdoctr.models.utilsrrr4r __all__rlistr$__annotations__boolr6intr r:r7r5rDs: 222222......++++++ " #   tCy    S d c >R    H  vv vvv v  v  vvvvvvr7