o Zh$@sdZddlmZmZmZmZmZddlmZddl m Z m Z ddl m Z mZddlmZmZmZmZmZmZddlmZmZgd Zgd Zd d ZGd dde ZdgZdS)z"Image processor class for Idefics.)CallableDictListOptionalUnion)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?cCsB|jdkr|S|d}td|jd}t||}|d}|S)NRGBZRGBA)rr)modeconvertrnewsizealpha_composite)imageZ image_rgba backgroundrrc/var/www/auris/lib/python3.10/site-packages/transformers/models/idefics/image_processing_idefics.pyconvert_to_rgb&s    r cs eZdZdZdgZ      dded eeee efd eeee efd eed e d eeefddffdd Z ddddddde j fded eedeeeefd eeee efd eeee efdeed ee d eedeeee fde fddZZS)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_valuesNrTp? image_size image_mean image_stdimage_num_channels do_rescalerescale_factorreturnc sRtjdi|||_||_|dur|nt|_|dur|nt|_||_||_ dS)Nr) super__init__r%r(IDEFICS_STANDARD_MEANr&IDEFICS_STANDARD_STDr'r)r*)selfr%r&r'r(r)r*kwargs __class__rrr-Ns  zIdeficsImageProcessor.__init__images transformreturn_tensorsc  s^|dur|nj}|dur|nj}durnjdur!nj|dur*|nj}dur3nj||ft|trGt|dkrGgSt |}t |sSt ddurpt s^t dddl} fdd|D}| |Sdd|D}dd|D}fd d|D}fd d|D}fd d|D}d d|D}td |i| dd }|S)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 installedcsg|]}|qSrr.0x)r5rr z4IdeficsImageProcessor.preprocess..cSg|]}t|qSr)r r7rrrr:r;cSr<r)rr7rrrr:r;csg|] }t|tjdqS))Zresample)r rZBICUBICr7)rrrr:csg|] }j|dqS))rscale)Zrescale)r8r)r*r0rrr:scsg|] }j|dqS))meanZstd) normalizer7)r&r'r0rrr:r=cSsg|]}t|tjqSr)r r ZFIRSTr7rrrr:sr")dataZ tensor_type)r%r(r&r'r)r* isinstancelistlenrr ValueErrorr ImportErrortorchstackr ) r0r4r(r%r&r'r5r)r*r6r1rGr)r&r'r*r0rr5r preprocessas8.  z IdeficsImageProcessor.preprocess)r#NNrTr$)__name__ __module__ __qualname____doc__Zmodel_input_namesintrrfloatrboolr-rZPYTORCHrrstrrrI __classcell__rrr2rr!3sj     r!N)rMtypingrrrrrZPILrZimage_processing_utilsr r Zimage_transformsr r Z image_utilsr rrrrrutilsrrr.r/r r!__all__rrrrs