§
    ·¯¸jÎW  ã            C       óp  — d Z ddlZddlmZmZmZ ddlZddlmZ ddl	m
Z
mZmZ ddlmZmZmZ ddlmZmZmZmZmZmZmZmZmZ ddlmZmZmZ dd	lm Z  d
de
eddfdee!ee!e!f         f         de"dee#df         dee#df         de$de$fd„Z%d
dddddddddddde
eddddddddddddfdee!ee!e!f         f         deee#e#f                  deee#e#f                  d ee"         d!e#d"e#d#ee#ee#df         f         d$ee#         d%e$d&e#d'e#d(ee"         de"dee#df         dee#df         d)e#d*e"d+e!d,e!de$de$d-e$d.e$d/ee!ee!e!f         f         d0e!d1e$d2e$f6d3„Z&d
dddde
eddddddddfdee!ee!e!f         f         d4ee#         d5ee"         d6ee!         de"dee#df         dee#df         de$de$d.e$d/ee!ee!e!f         f         d0e!d1e$d2e$d7e$fd8„Z'd
ddddddddddddde
edddddddddddddddddf!d9ee!ee!e!f         ee!e!e!f         f         d:e$d;e$d ee"         deee#e#f                  deee#e#f                  d!e#d"e#d#ee#ee#df         f         d$ee#         d&e#d'e#d(ee"         de"dee#df         dee#df         d)e#d*e"d+e!d,e!d4ee#         d5ee"         d6ee!         d<e$de$de$d-e$d.e$d/ee!ee!e!f         f         d0e!d1e$d2e$d7e$fBd=„Z(dS )>zŸ Transforms Factory
Factory methods for building image transforms for use with TIMM (PyTorch Image Models)

Hacked together by / Copyright 2019, Ross Wightman
é    N)ÚOptionalÚTupleÚUnion)Ú
transforms)ÚIMAGENET_DEFAULT_MEANÚIMAGENET_DEFAULT_STDÚDEFAULT_CROP_PCT)Úrand_augment_transformÚaugment_and_mix_transformÚauto_augment_transform)	Ústr_to_interp_modeÚstr_to_pil_interpÚ!RandomResizedCropAndInterpolationÚResizeKeepRatioÚCenterCropOrPadÚRandomCropOrPadÚ
TrimBorderÚMaybeToTensorÚMaybePILToTensor)ÚRandomResizedCropToSequenceÚResizeToSequenceÚPatchify)ÚRandomErasingéà   ÚbilinearFTÚimg_sizeÚinterpolationÚmean.ÚstdÚuse_prefetcherÚ	normalizec                 ó’  — |dk    rd}t          j        | t          |¦  «        ¬¦  «        t          j        | ¦  «        g}|r|t	          ¦   «         gz  }n`|s|t	          ¦   «         gz  }nK|t          ¦   «         t          j        t          j        |¦  «        t          j        |¦  «        ¬¦  «        gz  }t          j	        |¦  «        S )a¸   No-augmentation image transforms for training.

    Args:
        img_size: Target image size.
        interpolation: Image interpolation mode.
        mean: Image normalization mean.
        std: Image normalization standard deviation.
        use_prefetcher: Prefetcher enabled. Do not convert image to tensor or normalize.
        normalize: Normalization tensor output w/ provided mean/std (if prefetcher not used).

    Returns:

    Úrandomr   ©r   ©r   r   )
r   ÚResizer   Ú
CenterCropr   r   Ú	NormalizeÚtorchÚtensorÚCompose)r   r   r   r   r    r!   Útfls          úl/root/private/project/pallet_detection_v2/.venv/lib/python3.11/site-packages/timm/data/transforms_factory.pyÚtransforms_noaug_trainr.      sÛ   € ð* ˜Ò Ð à"ˆåÔ˜(Õ2DÀ]Ñ2SÔ2SÐTÑTÔTÝÔ˜hÑ'Ô'ð€Cð ð 
àÕ Ñ"Ô"Ð#Ñ#ˆˆØð 

àÕ Ñ"Ô"Ð#Ñ#ˆˆàÝ‰OŒOÝÔ Ý”\ $Ñ'Ô'Ý”L Ñ%Ô%ðñ ô ð
ñ 	
ˆõ Ô˜cÑ"Ô"Ð"ó    ç      à?ç        gš™™™™™Ù?r#   Úconsté   é   i@  ÚscaleÚratioÚtrain_crop_modeÚhflipÚvflipÚcolor_jitterÚcolor_jitter_probÚforce_color_jitterÚgrayscale_probÚgaussian_blur_probÚauto_augmentÚre_probÚre_modeÚre_countÚre_num_splitsÚseparateÚnaflexÚ
patch_sizeÚmax_seq_lenÚpatchifyÚpatchify_channels_lastc           
      ó  — |pd}|dv sJ ‚g }|r;t          |pd¦  «        }t          |pd¦  «        }|t          |||||¬¦  «        gz  }n¡|dv rdt          |pd¦  «        }t          |pd¦  «        }|t          | |d	|d
d	|¬¦  «        |dk    rt          | d¬¦  «        nt	          | d¬¦  «        gz  }n9t          |pd¦  «        }t          |pd¦  «        }|t          | |||¬¦  «        gz  }|dk    r|t          j        |¬¦  «        gz  }|dk    r|t          j        |¬¦  «        gz  }g }d}|�rt          |t          ¦  «        sJ ‚|pd|v  }t          | t           t          f¦  «        rt          | ¦  «        }n| }t          t          |dz  ¦  «        t          d„ |D ¦   «         ¦  «        ¬¦  «        }|r|dk    rt          |¦  «        |d<   |                     d¦  «        r|t#          ||¦  «        gz  }nC|                     d¦  «        rd|d<   |t%          ||¦  «        gz  }n|t'          ||¦  «        gz  }|��|st          |t          t           f¦  «        rt)          |¦  «        dv sJ ‚nt+          |¦  «        fdz  }|�(|t          j        t          j        |Ž g|¬¦  «        gz  }n|t          j        |Ž gz  }|	r|t          j        |	¬¦  «        gz  }|
r.|t          j        t          j        d ¬!¦  «        g|
¬¦  «        gz  }g } |r| t5          ¦   «         gz  } n~|s| t5          ¦   «         gz  } ni| t7          ¦   «         t          j        t;          j        |¦  «        t;          j        |¦  «        ¬"¦  «        gz  } |dk    r| t?          ||||d#¬$¦  «        gz  } |r| tA          ||¬%¦  «        gz  } |r;t          j!        |¦  «        t          j!        |¦  «        t          j!        | ¦  «        fS t          j!        ||z   | z   ¦  «        S )&a¬   ImageNet-oriented image transforms for training.

    Args:
        img_size: Target image size.
        train_crop_mode: Training random crop mode ('rrc', 'rkrc', 'rkrr').
        scale: Random resize scale range (crop area, < 1.0 => zoom in).
        ratio: Random aspect ratio range (crop ratio for RRC, ratio adjustment factor for RKR).
        hflip: Horizontal flip probability.
        vflip: Vertical flip probability.
        color_jitter: Random color jitter component factors (brightness, contrast, saturation, hue).
            Scalar is applied as (scalar,) * 3 (no hue).
        color_jitter_prob: Apply color jitter with this probability if not None (for SimlCLR-like aug).
        force_color_jitter: Force color jitter where it is normally disabled (ie with RandAugment on).
        grayscale_prob: Probability of converting image to grayscale (for SimCLR-like aug).
        gaussian_blur_prob: Probability of applying gaussian blur (for SimCLR-like aug).
        auto_augment: Auto augment configuration string (see auto_augment.py).
        interpolation: Image interpolation mode.
        mean: Image normalization mean.
        std: Image normalization standard deviation.
        re_prob: Random erasing probability.
        re_mode: Random erasing fill mode.
        re_count: Number of random erasing regions.
        re_num_splits: Control split of random erasing across batch size.
        use_prefetcher: Prefetcher enabled. Do not convert image to tensor or normalize.
        normalize: Normalize tensor output w/ provided mean/std (if prefetcher not used).
        separate: Output transforms in 3-stage tuple.
        naflex: Enable NaFlex mode, sequence constrained patch output
        patch_size: Patch size for NaFlex mode.
        max_seq_len: Max sequence length for NaFlex mode.

    Returns:
        If separate==True, the transforms are returned as a tuple of 3 separate transforms
          for use in a mixing dataset that passes
            * all data through the first (primary) transform, called the 'clean' data
            * a portion of the data through the secondary transform
            * normalizes and converts the branches above with the third, final transform
    Úrrc>   rK   ÚrkrcÚrkrr)g{®Gáz´?ç      ð?)g      è?gUUUUUUõ?)rF   rG   r5   r6   r   )rL   rM   )gš™™™™™é?rN   )gÍÌÌÌÌÌì?grÇqÇñ?r0   T)r   Úrandom_scale_probÚrandom_scale_rangeÚrandom_scale_areaÚrandom_aspect_probÚrandom_aspect_rangerL   Úreflect)Úpadding_mode)r5   r6   r   r1   )ÚpFÚ3agÍÌÌÌÌÌÜ?c           	      óN   — g | ]"}t          d t          d |z  ¦  «        ¦  «        ‘Œ#S ©éÿ   )ÚminÚround)Ú.0Úxs     r-   ú
<listcomp>z-transforms_imagenet_train.<locals>.<listcomp>Â   s,   € ÐCÐCÐC¸�C ¥U¨3°©7¡^¤^Ñ4Ô4ÐCÐCÐCr/   )Útranslate_constÚimg_meanr#   r   ÚrandÚaugmixg333333Ó?Útranslate_pctN)é   é   re   é   )Úkernel_sizer%   Úcpu)ÚmodeÚ	max_countÚ
num_splitsÚdevice)rF   Úchannels_last)"Útupler   r   r   r   r   r   ÚRandomHorizontalFlipÚRandomVerticalFlipÚ
isinstanceÚstrÚlistr[   ÚdictÚintr   Ú
startswithr
   r   r   ÚlenÚfloatÚRandomApplyÚColorJitterÚRandomGrayscaleÚGaussianBlurr   r   r(   r)   r*   r   r   r+   )!r   r5   r6   r7   r8   r9   r:   r;   r<   r=   r>   r?   r   r   r   r@   rA   rB   rC   r    r!   rD   rE   rF   rG   rH   rI   Úprimary_tflÚsecondary_tflÚdisable_color_jitterÚimg_size_minÚ	aa_paramsÚ	final_tfls!                                    r-   Útransforms_imagenet_trainr„   A   sf  € ðD &Ð.¨€OØÐ5Ð5Ð5Ð5Ð5à€KØð 'Ý�eÐ*˜{Ñ+Ô+ˆÝ�eÐ1Ð1Ñ2Ô2ˆØÕ3Ø!Ø#ØØØ'ð
ñ 
ô 
ð ñ 	ˆˆð Ð.Ð.Ð.å˜%Ð. ;Ñ/Ô/ˆEÝ˜%Ð. ;Ñ/Ô/ˆEØÝØØ"/Ø&)Ø',Ø&*Ø'*Ø(-ðñ ô ð # fÒ,Ð,õ   °yÐAÑAÔAÐAå °yÐAÑAÔAðñ ˆKˆKõ ˜%Ð. ;Ñ/Ô/ˆEÝ˜%Ð5Ð#5Ñ6Ô6ˆEØÝ1ØØØØ"/ð	ñ ô ðñ ˆKð ˆr‚z€zØ�
Ô7¸%Ð@Ñ@Ô@ÐAÑAˆØˆr‚z€zØ�
Ô5¸Ð>Ñ>Ô>Ð?Ñ?ˆà€MØ ÐØñ OÝ˜,­Ñ,Ô,Ð,Ð,Ð,ð %7Ð$N¸$À,Ð:NÐOÐÝ�h¥­ Ñ.Ô.ð 	$Ý˜x™=œ=ˆLˆLà#ˆLÝÝ ¨tÑ 3Ñ4Ô4ÝÐCÐC¸dÐCÑCÔCÑDÔDð
ñ 
ô 
ˆ	ð ð 	J˜]¨hÒ6Ð6Ý):¸=Ñ)IÔ)IˆI�oÑ&Ø×"Ò" 6Ñ*Ô*ð 	OØÕ4°\À9ÑMÔMÐNÑNˆMˆMØ×$Ò$ XÑ.Ô.ð 	OØ),ˆI�oÑ&ØÕ7¸ÀiÑPÔPÐQÑQˆMˆMàÕ4°\À9ÑMÔMÐNÑNˆMàÐÐ(<Ðå�l¥T­5 MÑ2Ô2ð 	6õ �|Ñ$Ô$¨Ð.Ð.Ð.Ð.Ð.õ " ,Ñ/Ô/Ð1°AÑ5ˆLØÐ(ØÝÔ&Ý"Ô.°Ð=ð(ð (ðñ ô ðñ ˆMˆMð �jÔ4°lÐCÐDÑDˆMàð HØ�*Ô4°~ÐFÑFÔFÐGÑGˆàð 
ØÝÔ"ÝÔ+¸Ð;Ñ;Ô;ð$ð %ðñ ô ð
ñ 	
ˆð €IØð àÕ&Ñ(Ô(Ð)Ñ)ˆ	ˆ	Øð àÕ&Ñ(Ô(Ð)Ñ)ˆ	ˆ	àÝ‰OŒOÝÔ Ý”\ $Ñ'Ô'Ý”L Ñ%Ô%ðñ ô ð
ñ 	
ˆ	ð �RŠ<ˆ<ØÝØØ Ø&Ø,Ø ðñ ô ðñ ˆIð ð ]Ø•h¨*ÐDZÐ[Ñ[Ô[Ð\Ñ\ˆ	àð KÝÔ! +Ñ.Ô.µ
Ô0BÀ=Ñ0QÔ0QÕS]ÔSeÐfoÑSpÔSpÐpÐpåÔ! +°Ñ"=À	Ñ"IÑJÔJÐJr/   Úcrop_pctÚ	crop_modeÚcrop_border_pixelsÚpatchify_flattenc                 óR  ‡— ‰pt           Št          | t          t          f¦  «        r1t	          | ¦  «        dk    sJ ‚t          ˆfd„| D ¦   «         ¦  «        }nt          j        | ‰z  ¦  «        }||f}g }|r|t          |¦  «        gz  }|	r|t          |
||¬¦  «        gz  }nå|dk    r;|t          j
        |t          |¦  «        ¬¦  «        t          j        | ¦  «        gz  }n¤|dk    r3d„ |D ¦   «         }|t          ||d¬	¦  «        t          | |¬
¦  «        gz  }nk|d         |d         k    r.|t          j
        |d         t          |¦  «        ¬¦  «        gz  }n|t          |¦  «        gz  }|t          j        | ¦  «        gz  }|r|t          ¦   «         gz  }n`|s|t          ¦   «         gz  }nK|t!          ¦   «         t          j        t%          j        |¦  «        t%          j        |¦  «        ¬¦  «        gz  }|r|t)          |
||¬¦  «        gz  }t          j        |¦  «        S )aÏ   ImageNet-oriented image transform for evaluation and inference.

    Args:
        img_size: Target image size.
        crop_pct: Crop percentage. Defaults to 0.875 when None.
        crop_mode: Crop mode. One of ['squash', 'border', 'center']. Defaults to 'center' when None.
        crop_border_pixels: Trim a border of specified # pixels around edge of original image.
        interpolation: Image interpolation mode.
        mean: Image normalization mean.
        std: Image normalization standard deviation.
        use_prefetcher: Prefetcher enabled. Do not convert image to tensor or normalize.
        normalize: Normalize tensor output w/ provided mean/std (if prefetcher not used).
        naflex: Enable NaFlex mode, sequence constrained patch output
        patch_size: Patch size for NaFlex mode.
        max_seq_len: Max sequence length for NaFlex mode.
        patchify: Patchify the output instead of relying on prefetcher
        patchify_channels_last: Use channels-last layout within each patch.
        patchify_flatten: Flatten each patch into a vector. Disable for variable
            patch-size interpolation, which requires spatial patch dimensions.

    Returns:
        Composed transform pipeline
    é   c                 ó>   •— g | ]}t          j        |‰z  ¦  «        ‘ŒS © )ÚmathÚfloor)r]   r^   r…   s     €r-   r_   z,transforms_imagenet_eval.<locals>.<listcomp>=  s'   ø€ ÐGÐGÐG¸�DœJ q¨8¡|Ñ4Ô4ÐGÐGÐGr/   )rF   rG   r   Úsquashr$   Úborderc                 ó2   — g | ]}t          d |z  ¦  «        ‘ŒS rY   )r\   )r]   Úvs     r-   r_   z,transforms_imagenet_eval.<locals>.<listcomp>X  s"   € Ð1Ð1Ð1 q•E˜# ™'‘N”NÐ1Ð1Ð1r/   rN   )r   Úlongest)Úfillr   r3   r%   )rF   Úflatten_patchesrn   )r	   rr   ro   rt   rx   r�   rŽ   r   r   r   r&   r   r'   r   r   r   r   r(   r)   r*   r   r+   )r   r…   r†   r‡   r   r   r   r    r!   rE   rF   rG   rH   rI   rˆ   Ú
scale_sizer,   r”   s    `                r-   Útransforms_imagenet_evalr—     s´  ø€ ðP Ð+Õ+€Hå�(�U¥D˜MÑ*Ô*ð .Ý�8‰}Œ} Ò!Ð!Ð!Ð!ÝÐGÐGÐGÐG¸hÐGÑGÔGÑHÔHˆ
ˆ
å”Z ¨8Ñ 3Ñ4Ô4ˆ
Ø  *Ð-ˆ
à
€Càð 0Ø•
Ð-Ñ.Ô.Ð/Ñ/ˆàð !5ØÕ Ø!Ø#Ø'ð
ñ 
ô 
ð ñ 	ˆˆð ˜Ò Ð ð ÝÔ! *Õ<NÈ}Ñ<]Ô<]Ð^Ñ^Ô^ÝÔ% hÑ/Ô/ðñ ˆCˆCð ˜(Ò"Ð"ð 2Ð1¨DÐ1Ñ1Ô1ˆDØÝ 
¸-ÐQTÐUÑUÔUÝ ¨tÐ4Ñ4Ô4ðñ ˆCˆCð ˜!Œ} 
¨1¤Ò-Ð-àÝÔ% j°¤mÕCUÐVcÑCdÔCdÐeÑeÔeðñ ��ð
 �¨
Ñ3Ô3Ð4Ñ4�Ø•JÔ)¨(Ñ3Ô3Ð4Ñ4ˆCàð 
àÕ Ñ"Ô"Ð#Ñ#ˆˆØð 

àÕ Ñ"Ô"Ð#Ñ#ˆˆàÝ‰OŒOÝÔ Ý”\ $Ñ'Ô'Ý”L Ñ%Ô%ðñ ô ð
ñ 	
ˆð ð Ø•Ø!Ø,Ø0ð
ñ 
ô 
ð ñ 	ˆõ Ô˜cÑ"Ô"Ð"r/   Ú
input_sizeÚis_trainingÚno_augÚtf_preprocessingc!                 óô  — t          | t          t          f¦  «        r| dd…         }!n| }!|r#|r!|r
J d¦   «         ‚ddlm}"  |"||!|¬¦  «        }#n©|r$|r"|r
J d¦   «         ‚t          |!|||||¬¦  «        }#nƒ|rWt          |!fi d	|“d
|“d|“d|“d|“d|“d|	“d|
“d|“d|“d|“d|“d|“d|“d|“d|“d|“d|“d|“d|“d|“d|“d|“d |“d!|“Ž}#n*|r
J d"¦   «         ‚t          |!|||||||||||||| ¬#¦  «        }#|#S )$al	  

    Args:
        input_size: Target input size (channels, height, width) tuple or size scalar.
        is_training: Return training (random) transforms.
        no_aug: Disable augmentation for training (useful for debug).
        train_crop_mode: Training random crop mode ('rrc', 'rkrc', 'rkrr').
        scale: Random resize scale range (crop area, < 1.0 => zoom in).
        ratio: Random aspect ratio range (crop ratio for RRC, ratio adjustment factor for RKR).
        hflip: Horizontal flip probability.
        vflip: Vertical flip probability.
        color_jitter: Random color jitter component factors (brightness, contrast, saturation, hue).
            Scalar is applied as (scalar,) * 3 (no hue).
        color_jitter_prob: Apply color jitter with this probability if not None (for SimlCLR-like aug).
        grayscale_prob: Probability of converting image to grayscale (for SimCLR-like aug).
        gaussian_blur_prob: Probability of applying gaussian blur (for SimCLR-like aug).
        auto_augment: Auto augment configuration string (see auto_augment.py).
        interpolation: Image interpolation mode.
        mean: Image normalization mean.
        std: Image normalization standard deviation.
        re_prob: Random erasing probability.
        re_mode: Random erasing fill mode.
        re_count: Number of random erasing regions.
        re_num_splits: Control split of random erasing across batch size.
        crop_pct: Inference crop percentage (output size / resize size).
        crop_mode: Inference crop mode. One of ['squash', 'border', 'center']. Defaults to 'center' when None.
        crop_border_pixels: Inference crop border of specified # pixels around edge of original image.
        tf_preprocessing: Use TF 1.0 inference preprocessing for testing model ports
        use_prefetcher: Pre-fetcher enabled. Do not convert image to tensor or normalize.
        normalize: Normalization tensor output w/ provided mean/std (if prefetcher not used).
        separate: Output transforms in 3-stage tuple.
        patchify: Patchify the output instead of relying on prefetcher.
        patchify_channels_last: Use channels-last layout within each patch.
        patchify_flatten: Flatten eval patches into vectors. Disable when the
            model needs their spatial dimensions for patch-size interpolation.

    Returns:
        Composed transforms or tuple thereof
    éþÿÿÿNz6Separate transforms not supported for TF preprocessingr   )ÚTfPreprocessTransform)r™   Úsizer   z-Cannot perform split augmentation with no_aug)r   r   r   r    r!   r7   r5   r6   r8   r9   r:   r;   r=   r>   r?   r   r   r   r@   rA   rB   rC   r    r!   rD   rE   rF   rG   rH   rI   z>Separate transforms not supported for validation preprocessing)r   r   r   r…   r†   r‡   r    r!   rE   rF   rG   rH   rI   rˆ   )rr   ro   rt   Útimm.data.tf_preprocessingrž   r.   r„   r—   )$r˜   r™   rš   r7   r5   r6   r8   r9   r:   r;   r=   r>   r?   r   r   r   r@   rA   rB   rC   r…   r†   r‡   r›   r    r!   rD   rE   rF   rG   rH   rI   rˆ   r   rž   Ú	transforms$                                       r-   Úcreate_transformr¢   ƒ  s„  € õT �*�u¥d˜mÑ,Ô,ð Ø˜b˜c˜c”?ˆˆàˆàð B˜Nð BØÐUÐUÐUÑUÔUˆ|ØDÐDÐDÐDÐDÐDØ)Ð)Ø#ØØ'ð
ñ 
ô 
ˆ	ˆ	ð ð 9	˜6ð 9	ØÐPÐPÐ!PÑPÔP�<Ý.ØØ+ØØØ-Ø#ðñ ô ˆIˆIð ð /	Ý1Øðð ð à / ðð �eðð �eð	ð
 �eðð �eðð *˜\ðð #4Ð"3ðð  .˜~ðð $6Ð#5ðð *˜\ðð ,˜mðð �Tðð �Cðð  ˜ðð   ˜ð!ð" "˜ð#ð$ ,˜mð%ð&  .˜~ð'ð( $˜)ð)ð* "˜ð+ð, �vð-ð. &˜:ð/ð0 (˜Kð1ð2 "˜ð3ð4 (>Ð'=ð5ˆIˆIð:  ÐaÐaÐ!aÑaÔa�<Ý0ØØ+ØØØ!Ø#Ø#5Ø-Ø#ØØ%Ø'Ø!Ø'=Ø!1ðñ ô ˆIð$ Ðr/   ))Ú__doc__r�   Útypingr   r   r   r)   Útorchvisionr   Útimm.data.constantsr   r   r	   Útimm.data.auto_augmentr
   r   r   Útimm.data.transformsr   r   r   r   r   r   r   r   r   Útimm.data.naflex_transformsr   r   r   Útimm.data.random_erasingr   rv   rs   ry   Úboolr.   r„   r—   r¢   rŒ   r/   r-   ú<module>r¬      s­  ððð ð
 €€€Ø )Ð )Ð )Ð )Ð )Ð )Ð )Ð )Ð )Ð )à €€€Ø "Ð "Ð "Ð "Ð "Ð "à ]Ð ]Ð ]Ð ]Ð ]Ð ]Ð ]Ð ]Ð ]Ð ]Ø lÐ lÐ lÐ lÐ lÐ lÐ lÐ lÐ lÐ lðcð cð cð cð cð cð cð cð cð cð cð cð cð cð cð cð cð cð cð cð cð cà _Ð _Ð _Ð _Ð _Ð _Ð _Ð _Ð _Ð _Ø 2Ð 2Ð 2Ð 2Ð 2Ð 2ð 14Ø'Ø"7Ø!5Ø$Øð*#ð *#Ø˜˜U 3¨ 8œ_Ð,Ô-ð*#àð*#ð �E˜3�JÔð*#ð �5˜#�:Ôð	*#ð
 ð*#ð ð*#ð *#ð *#ð *#ð\ 14Ø/3Ø/3Ø)-ØØØ8;Ø-1Ø#(Ø "Ø$&Ø&*Ø%Ø"7Ø!5ØØØØØ$ØØØØ24ØØØ'+ð7MKð MKØ˜˜U 3¨ 8œ_Ð,Ô-ðMKà˜˜e U˜lÔ+Ô,ðMKð ˜˜e U˜lÔ+Ô,ðMKð " #œð	MKð
 ðMKð ðMKð ˜E 5¨°¨Ô#4Ð4Ô5ðMKð $ Eœ?ðMKð !ðMKð ðMKð "ðMKð ˜s”mðMKð ðMKð �E˜3�JÔðMKð �5˜#�:ÔðMKð  ð!MKð" ð#MKð$ ð%MKð& ð'MKð( ð)MKð* ð+MKð, ð-MKð. ð/MKð0 ˜#˜u S¨# XœÐ.Ô/ð1MKð2 ð3MKð4 ð5MKð6 !%ð7MKð MKð MKð MKðb 14Ø$(Ø#'Ø,0Ø'Ø"7Ø!5Ø$ØØØ24ØØØ'+Ø!%ðo#ð o#Ø˜˜U 3¨ 8œ_Ð,Ô-ðo#à˜5”/ðo#ð ˜C”=ðo#ð % SœMð	o#ð
 ðo#ð �E˜3�JÔðo#ð �5˜#�:Ôðo#ð ðo#ð ðo#ð ðo#ð ˜#˜u S¨# XœÐ.Ô/ðo#ð ðo#ð ðo#ð !%ðo#ð ðo#ð o#ð o#ð o#ðf ILØ!ØØ)-Ø/3Ø/3ØØØ8;Ø-1Ø "Ø$&Ø&*Ø'Ø"7Ø!5ØØØØØ$(Ø#'Ø,0Ø!&Ø$ØØØØ24ØØØ'+Ø!%ðCSð SØ˜#˜u S¨# Xœ°°c¸3À°mÔ0DÐDÔEðSàðSð ðSð " #œð	Sð
 ˜˜e U˜lÔ+Ô,ðSð ˜˜e U˜lÔ+Ô,ðSð ðSð ðSð ˜E 5¨°¨Ô#4Ð4Ô5ðSð $ Eœ?ðSð ðSð "ðSð ˜s”mðSð ðSð �E˜3�JÔðSð  �5˜#�:Ôð!Sð" ð#Sð$ ð%Sð& ð'Sð( ð)Sð* ˜5”/ð+Sð, ˜C”=ð-Sð. % SœMð/Sð0 ð1Sð2 ð3Sð4 ð5Sð6 ð7Sð8 ð9Sð: ˜#˜u S¨# XœÐ.Ô/ð;Sð< ð=Sð> ð?Sð@ !%ðASðB ðCSð Sð Sð Sð Sð Sr/   