rh$dZddlmZmZmZddlmZddlmZm Z ddl m Z m Z ddl mZmZmZmZmZmZddlmZmZgd Zgd Zd ZGd d eZd gZy)z"Image processor class for Idefics.)CallableOptionalUnion)Image)BaseImageProcessor BatchFeature)resizeto_channel_dimension_format)ChannelDimension ImageInputPILImageResamplingmake_list_of_imagesto_numpy_array valid_images) TensorTypeis_torch_available)g3<4'?gwg M?gy{ ?)gB91?gwt.?g U?c|jdk(r|S|jd}tjd|jd}tj ||}|jd}|S)NRGBRGBA)rr)modeconvertrnewsizealpha_composite)image image_rgba backgroundrs /var/www/html/ai-insurance-compliance-backend/venv/lib/python3.12/site-packages/transformers/models/idefics/image_processing_idefics.pyconvert_to_rgbr!&sa zzU v&J6:??ODJ++J CO%--e4O cJeZdZdZdgZ ddedeeee efdeeee efdeed e d eeefd dffd Z ddddddde jfd edeedeeeefdeeee efdeeee efdeed ee d eedeeee fd e fdZxZS)IdeficsImageProcessora 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. do_rescale (`bool`, *optional*, defaults to `True`): Whether to rescale the image by the specified scale `rescale_factor`. Can be overridden by `do_rescale` in the `preprocess` method. rescale_factor (`int` or `float`, *optional*, defaults to `1/255`): Scale factor to use if rescaling the image. Can be overridden by `rescale_factor` in the `preprocess` method. pixel_valuesNr image_size image_mean image_stdimage_num_channels do_rescalerescale_factorreturnc t|di|||_||_||nt|_||nt |_||_||_ y)N) super__init__r&r)IDEFICS_STANDARD_MEANr'IDEFICS_STANDARD_STDr(r*r+) selfr&r'r(r)r*r+kwargs __class__s r r0zIdeficsImageProcessor.__init__NsQ "6"$"4(2(>*DY&/&;AU$,r"images transformreturn_tensorsc  ||n |j}||n |j}||n |j}||n |j}||n |j}||n |j }||f} t |trt|dk(rgSt|}t|s td|?ts tdddl} |D cgc] } ||  }} | j|S|D cgc] } t!| }} |D cgc] } t#| }} |D cgc]} t%| | t&j( }} |Dcgc]}|j+||}}|D cgc]} |j-| ||}} |D cgc]} t/| t0j2}} t5d|i| d}|Scc} wcc} wcc} wcc} wcc}wcc} wcc} w) a| 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 do_rescale (`bool`, *optional*, defaults to `True`): Whether to rescale the image by the specified scale `rescale_factor`. Can be overridden by `do_rescale` in the `preprocess` method. rescale_factor (`int` or `float`, *optional*, defaults to `1/255`): Scale factor to use if rescaling the image. Can be overridden by `rescale_factor` in the `preprocess` method. 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)resample)rscale)meanstdr%)data tensor_type)r&r)r'r(r*r+ isinstancelistlenrr ValueErrorr ImportErrortorchstackr!rr rBICUBICrescale normalizer r FIRSTr )r3r6r)r&r'r(r7r*r+r8r4rrExrs r preprocessz IdeficsImageProcessor.preprocessas\$.#9Zt 3E3Q/W[WnWn#-#9Zt !*!6IDNN #-#9Zt +9+E4K^K^J' fd #F q(8I$V,F#:   %'!"RSS ,23qil3F3;;v& &.44.#44-34.#44PVW1&D+=+E+EFWWOUVe$,,U.,AVVMST$...CTTRXYQ-a1A1G1GHYYNF#;XYgh 454WVTYs*F<+GG#G G$G!G)NNrTgp?)__name__ __module__ __qualname____doc__model_input_namesintrrfloatrAboolr0rPYTORCHr dictstrrrL __classcell__)r5s@r r$r$3s0((:>9=,-,3--U5$u+#567-E%e"456 - %SM -  -c5j)- -,-./3:>9=(,%)*.;E;M;M[[%SM[T#s(^, [ U5$u+#567 [ E%e"456 [H%[TN[![!sJ!78[ [r"r$N)rQtypingrrrPILrimage_processing_utilsrr image_transformsr r image_utilsr r rrrrutilsrrr1r2r!r$__all__r.r"r rasU),,FC4<; I.IX # #r"