Ë ©sgTãó¬—dZddlmZmZmZmZmZddlmZddl m Z m Z ddl m Z mZddlmZmZmZmZmZmZddlmZmZgd ¢Zgd ¢Zd „ZGd „d e «Zy)z"Image processor class for Idefics.é)ÚCallableÚDictÚListÚOptionalÚUnion)ÚImageé)ÚBaseImageProcessorÚ BatchFeature)ÚresizeÚto_channel_dimension_format)ÚChannelDimensionÚ ImageInputÚPILImageResamplingÚmake_list_of_imagesÚto_numpy_arrayÚ valid_images)Ú TensorTypeÚis_torch_available)g3<Í4'ÐÞ?gwgí¶ MÝ?gy{Îå Ú?)g‡Bô91Ñ?g•wÝt.¹Ð?gÝ U¦Ñ?cóÚ—|jdk(r|S|jd«}tjd|jd«}tj ||«}|jd«}|S)NÚRGBÚRGBA)éÿrr)ÚmodeÚconvertrÚnewÚsizeÚalpha_composite)ÚimageÚ image_rgbaÚ backgroundrs úg/var/www/html/venv/lib/python3.12/site-packages/transformers/models/idefics/image_processing_idefics.pyÚconvert_to_rgbr#&sa€ð ‡zzUÒ؈ à—‘˜vÓ&€JÜ—‘˜6 :§?¡?°OÓD€JÜ×+Ñ+¨J¸ ÓC€OØ%×-Ñ-¨eÓ4€OØ Ðócó‡—eZdZdZdgZ ddedeeee efdeeee efdeed df ˆfd „ Z ddddde jfd e deedeeeefdeeee efdeeee efd ed eeee fd e fd„ZˆxZS)ÚIdeficsImageProcessora1 Constructs a Idefics image processor. Args: image_size (`int`, *optional*, defaults to 224): Resize to image size image_mean (`float` or `List[float]`, *optional*, defaults to `IDEFICS_STANDARD_MEAN`): Mean to use if normalizing the image. This is a float or list of floats the length of the number of channels in the image. Can be overridden by the `image_mean` parameter in the `preprocess` method. Can be overridden by the `image_mean` parameter in the `preprocess` method. image_std (`float` or `List[float]`, *optional*, defaults to `IDEFICS_STANDARD_STD`): Standard deviation to use if normalizing the image. This is a float or list of floats the length of the number of channels in the image. Can be overridden by the `image_std` parameter in the `preprocess` method. Can be overridden by the `image_std` parameter in the `preprocess` method. image_num_channels (`int`, *optional*, defaults to 3): Number of image channels. Ú pixel_valuesNr Ú image_sizeÚ image_meanÚ image_stdÚimage_num_channelsÚreturnc ó\•—t‰|di|¤Ž||_||_||_||_y)N©)ÚsuperÚ__init__r(r+r)r*)Úselfr(r)r*r+ÚkwargsÚ __class__s €r"r0zIdeficsImageProcessor.__init__Hs2ø€ô ‰ÑÑ"˜6Ò"à$ˆŒØ"4ˆÔØ$ˆŒØ"ˆr$ÚimagesÚ transformÚreturn_tensorsc ó~—||n |j}||n |j}||n |j}||n |j}||f} t |t «rt |«dk(rgSt|«}t|«s td«‚|?t«s td«‚ddl } |D cgc] } || «‘Œ }} | j|«S|D cgc] } t| «‘Œ}} |D cgc] } t| «‘Œ}} |D cgc]} t!| | t"j$¬«‘Œ }} |D cgc]} |j'| d¬«‘Œ}} |D cgc]} |j)| ||¬«‘Œ}} |D cgc]} t+| t,j.«‘Œ}} t1d |i|¬ «d }|Scc} wcc} wcc} wcc} wcc} wcc} wcc} w) a¹ Preprocess a batch of images. Args: images (`ImageInput`): A list of images to preprocess. image_size (`int`, *optional*, defaults to `self.image_size`): Resize to image size image_num_channels (`int`, *optional*, defaults to `self.image_num_channels`): Number of image channels. image_mean (`float` or `List[float]`, *optional*, defaults to `IDEFICS_STANDARD_MEAN`): Mean to use if normalizing the image. This is a float or list of floats the length of the number of channels in the image. Can be overridden by the `image_mean` parameter in the `preprocess` method. Can be overridden by the `image_mean` parameter in the `preprocess` method. image_std (`float` or `List[float]`, *optional*, defaults to `IDEFICS_STANDARD_STD`): Standard deviation to use if normalizing the image. This is a float or list of floats the length of the number of channels in the image. Can be overridden by the `image_std` parameter in the `preprocess` method. Can be overridden by the `image_std` parameter in the `preprocess` method. transform (`Callable`, *optional*, defaults to `None`): A custom transform function that accepts a single image can be passed for training. For example, `torchvision.Compose` can be used to compose multiple transforms. If `None` - an inference mode is assumed - and then a preset of inference-specific transforms will be applied to the images Returns: a PyTorch tensor of the processed images NrzkInvalid image type. Must be of type PIL.Image.Image, numpy.ndarray, torch.Tensor, tf.Tensor or jax.ndarray.z.To pass in `transform` torch must be installed)Úresamplegp?)rÚscale)ÚmeanÚstdr')ÚdataÚ tensor_type)r(r+r)r*Ú isinstanceÚlistÚlenrrÚ ValueErrorrÚ ImportErrorÚtorchÚstackr#rr rÚBICUBICÚrescaleÚ normalizer rÚFIRSTr ) r1r4r+r(r)r*r5r6r2rrCÚxrs r"Ú preprocessz IdeficsImageProcessor.preprocessWsÀðL$.Ð#9‘Z¸t¿¹ˆ Ø3EÐ3QÑ/ÐW[×WnÑWnÐØ#-Ð#9‘Z¸t¿¹ˆ Ø!*Ð!6‘I¸D¿N¹Nˆ ؘJÐ'ˆä fœdÔ #¬¨F« °qÒ(8؈Iä$ VÓ,ˆä˜FÔ#Üð:óð ð Ð Ü%Ô'Ü!Ð"RÓSÐSÛ à,2Ö3 q‘i •lÐ3ˆFÐ3Ø—;‘;˜vÓ&Ð &ð.4Ö4¨”. Õ#Ð4ˆÐ4à-3Ö4¨”. Õ#Ð4ˆÐ4ØPVÖWÈ1”&˜˜DÔ+=×+EÑ+EÖFÐWˆÐWØHNÖO¸u$—,‘, U°',Õ:ÐOˆÐOØMSÖTÈ$—.‘. ¨¸.ÕCÐTˆÐTØRXÖYÈQÔ-¨aÔ1A×1GÑ1GÕHÐYˆÐYÜ N°FÐ#;ÈÔXÐYgÑhˆàˆ ùò4ùò5ùâ4ùÚWùÚOùÚTùÚYs*Â%Fà F!Ã#F&Ã;#F+Ä$F0ÅF5Å%!F:)éàNNr )Ú__name__Ú __module__Ú __qualname__Ú__doc__Úmodel_input_namesÚintrrÚfloatrr0rÚPYTORCHrrÚstrrrJÚ __classcell__)r3s@r"r&r&3s;ø„ñð$(Ð(ÐðØ:>Ø9=Ø,-ñ #àð #ð˜U 5¨$¨u©+Ð#5Ñ6Ñ7ð #ð˜E %¨¨e©Ð"4Ñ5Ñ6ð #ð % S™Mð #ð õ #ð$-.Ø/3Ø:>Ø9=Ø"Ø;E×;MÑ;MñQàðQð% S™MðQð˜T # s (™^Ñ,ð Qð ˜U 5¨$¨u©+Ð#5Ñ6Ñ7ð Qð ˜E %¨¨e©Ð"4Ñ5Ñ6ð QððQð!  s¨J Ñ!7Ñ8ðQð ÷Qr$r&N)rOÚtypingrrrrrÚPILrÚimage_processing_utilsr r Úimage_transformsr r Ú image_utilsrrrrrrÚutilsrrÚIDEFICS_STANDARD_MEANÚIDEFICS_STANDARD_STDr#r&r.r$r"úr^sJðñ)ç8Õ8åçFßC÷÷÷4ò<ÐÚ;Ðò ôuÐ.õur$