+
    œ°³j4  ã                   óz   € R t ^ RIHt ^RIHt ^RIHt ^RIHt ]! RR7      ] ! R R]]4      4       4       t	R.t
R	# )
zHiera model configuration)Ústrict)ÚBackboneConfigMixin)ÚPreTrainedConfig)Úauto_docstringzfacebook/hiera-base-224)Ú
checkpointc                   ó¼   a a€ ] tR t^t oRtRtRR/t^`tRtRt	Rt
RtRtRtRtRt^tRtRtRtRt^tR	tR
tRtRtRtRtRtRtRtRtRt V 3R lt!R t"V3R lt#Rt$Vt%V ;t&# )ÚHieraConfigaá  
patch_stride (`list(int)`, *optional*, defaults to `[4, 4]`):
    The stride of the patch.
patch_padding (`list(int)`, *optional*, defaults to `[3, 3]`):
    The padding of the patch.
num_heads (`list(int)`, *optional*, defaults to `[1, 2, 4, 8]`):
    Number of attention heads in each layer of the Transformer encoder.
embed_dim_multiplier (`float`, *optional*, defaults to 2.0):
    The multiplier to the dimensionality of patch embedding in each layer of the Transformer encoder.
num_query_pool (`int`, *optional*, defaults to 3):
    The number of query pool stages.
query_stride (`list(int)`, *optional*, defaults to `[2, 2]`):
    The stride of the query pool.
masked_unit_size (`list(int)`, *optional*, defaults to `[8, 8]`):
    The size of the masked unit.
masked_unit_attention (`list(bool)`, *optional*, defaults to `[True, True, False, False]`):
    Whether to use masked unit attention in each layer of the Transformer encoder.
layer_norm_init (`float`, *optional*, defaults to 1.0):
    The initial weight value for layer normalization layers.
decoder_depth (`int`, *optional*):
    Depth of the decoder for MAE pretraining.
normalize_pixel_loss (`bool`, *optional*, defaults to `True`):
    Whether to normalize the pixel loss by the number of pixels.
mask_ratio (`float`, *optional*, defaults to 0.6):
    The ratio of masked tokens in the input.

Example:

```python
>>> from transformers import HieraConfig, HieraModel

>>> # Initializing a Hiera hiera-base-patch16-224 style configuration
>>> configuration = HieraConfig()

>>> # Initializing a model (with random weights) from the hiera-base-patch16-224 style configuration
>>> model = HieraModel(configuration)

>>> # Accessing the model configuration
>>> configuration = model.config
```ÚhieraÚnum_hidden_layersÚ
num_layersg      @g       @Tg        Úgelug{®Gáz”?g      ð?g�íµ ÷Æ°>Ng333333ã?c                ó¼  <€ \        V P                  V P                  \        V P                  4      ^,
          ,          ,          4      V n        R.\        ^\        V P                  4      ^,           4       Uu. uF  pRV 2NK
  	  up,           V n        V P                  VP                  RR4      VP                  RR4      R7       \        SV `,  ! R/ VB  R# u upi )é   ÚstemÚstageÚout_indicesNÚout_features)r   r   © )ÚintÚ	embed_dimÚembed_dim_multiplierÚlenÚdepthsÚhidden_sizeÚrangeÚstage_namesÚ"set_output_features_output_indicesÚpopÚsuperÚ__post_init__)ÚselfÚkwargsÚidxÚ	__class__s   &, €Ún/home/luca/screen-checking/.venv/lib/python3.14/site-packages/transformers/models/hiera/configuration_hiera.pyr   ÚHieraConfig.__post_init__b   s¼   ø€ ô ˜tŸ~™~°×0IÑ0IÌcÐRV×R]ÑR]ÓN^ÐabÕNbÕ0cÕcÓdˆÔØ"˜8ÄÀaÌÈTÏ[É[ÓIYÐ\]ÕI]Ô@^Ó&_Ñ@^¸¨¨s¨e£}Ñ@^Ñ&_Õ_ˆÔØ×/Ñ/ØŸ
™
 =°$Ó7ÀfÇjÁjÐQ_ÐaeÓFfð 	0ô 	
ô 	‰ÒÑ' Ô'ùò	 '`s   Á8Cc           
     óð  € V P                   ^ ,          V P                  ^ ,          \        V P                  4      ^,
          ,          ,          ^ 8w  dR   \	        RV P                   ^ ,           RV P                  ^ ,           R\        V P                  4      ^,
           R24      hV P
                  \        V P                  4      8¼  d0   \	        RV P
                   R\        V P                  4       R24      hR# )zOPart of `@strict`-powered validation. Validates the architecture of the config.zmasked_unit_size[0] (z() must be divisible by query_stride[0] (z/) raised to the power of the number of layers (Ú)znum_query_pool (z*) must be less than the number of layers (N)Úmasked_unit_sizeÚquery_strider   r   Ú
ValueErrorÚnum_query_pool)r    s   &r$   Úvalidate_architectureÚ!HieraConfig.validate_architecturel   sè   € à× Ñ  Õ# d×&7Ñ&7¸Õ&:¼sÀ4Ç;Á;Ó?OÐRSÕ?SÕ&TÕTÐXYÔYÜØ'¨×(=Ñ(=¸aÕ(@Ð'AÐAiÐjn×j{Ñj{Ð|}Õj~Ðið @@Ü@CÀDÇKÁKÓ@PÐSTÕ@TÐ?UÐUVðXóð ð
 ×Ñ¤# d§k¡kÓ"2Ô2ÜØ" 4×#6Ñ#6Ð"7Ð7aÔbeÐfj×fqÑfqÓbrÐasÐstÐuóð ñ 3ó    c                óº  <€ V ^8„  d   Qh/ S[ ;R&   S[S[ ,          S[S[ R3,          ,          ;R&   S[S[ ,          S[S[ R3,          ,          ;R&   S[S[ ,          S[S[ R3,          ,          ;R&   S[S[ ,          S[S[ R3,          ,          ;R&   S[;R&   S[S[ ,          S[S[ R3,          ,          ;R&   S[S[ ,          S[S[ R3,          ,          ;R	&   S[S[ ,          ;R
&   S[ ;R&   S[S[ ,          S[S[ R3,          ,          ;R&   S[S[ ,          S[S[ R3,          ,          ;R&   S[S[,          S[S[R3,          ,          ;R&   S[S[ ,          ;R&   S[ ;R&   S[;R&   S[;R&   S[;R&   S[;R&   S[ R,          ;R&   S[ R,          ;R&   S[ R,          ;R&   S[R,          ;R&   S[;R&   S[S[,          R,          ;R&   S[S[ ,          R,          ;R&   # )é   r   .Ú
image_sizeÚ
patch_sizeÚpatch_strideÚpatch_paddingÚ	mlp_ratior   Ú	num_headsr   r+   r)   r(   Úmasked_unit_attentionÚdrop_path_rateÚnum_channelsÚ
hidden_actÚinitializer_rangeÚlayer_norm_initÚlayer_norm_epsNÚdecoder_hidden_sizeÚdecoder_depthÚdecoder_num_headsÚnormalize_pixel_lossÚ
mask_ratioÚ_out_featuresÚ_out_indices)r   ÚlistÚtupleÚfloatÚboolÚstr)ÚformatÚ__classdict__s   "€r$   Ú__annotate__ÚHieraConfig.__annotate__   sO  ø‡ ‚ ñ` Ñña ñb ‘S•	™E¡# s (�OÕ+Ñ8ñc ñd ‘S•	™E¡# s (�OÕ+Ñ4ñe ñf ‘s•)™e¡C¨ H�oÕ-Ñ6ñg ñh ™•9™u¡S¨# X�Õ.Ñ7ñi ñj Ññk ñl ‘�I™™c 3˜h�Õ'Ñ7ñm ñn ‘C�y™5¡ c �?Õ*Ñ9ño ñp  ¡#�+Ñ+ñq ñr Ññs ñt ‘s•)™e¡C¨ H�oÕ-Ñ6ñu ñv ™3•i¡%©¨S¨¥/Õ1Ñ:ñw ñx  ¡�:©©d°C¨iÕ(8Õ8ÑUñy ñz ™C•KÑ%ñ{ ñ| Ññ} ñ~ Ññ ñ@ Ñ#ñA ñB Ñ ñC ñD Ñ ñE ñF ˜t�Ñ*ñG ñH ˜•:Ñ$ñI ñJ ˜T•zÑ(ñK ñL  �+Ñ,ñM ñN ÑñO ñP ™•9˜tÕ#Ñ*ñQ ñR ‘s•)˜dÕ"Ñ)òS r.   )r   r   )éà   rN   )é   rO   )é   rP   )é   rQ   )r0   rQ   é   rQ   )r   r0   rP   é   )r0   r0   )rS   rS   )TTFF)'Ú__name__Ú
__module__Ú__qualname__Ú__firstlineno__Ú__doc__Ú
model_typeÚattribute_mapr   r1   r2   r3   r4   r5   r   r6   r   r+   r)   r(   r7   r8   r9   r:   r;   r<   r=   r>   r?   r@   rA   rB   rC   rD   r   r,   Ú__annotate_func__Ú__static_attributes__Ú__classdictcell__Ú__classcell__)r#   rK   s   @@r$   r   r      s·   ù‡ € ñ'ðR €Jà(¨,Ð7€Mà€IØ.8€JØ.4€JØ06€LØ17€MØ€IØ*7€FØ-9€IØ(+ÐØ€NØ06€LØ4:ÐØ;UÐØ"%€NØ€LØ€JØ#ÐØ €OØ €NØ&*ÐØ $€MØ$(ÐØ(,ÐØ€JØ&*€MØ%)€Lõ(ò÷k … r.   r   N)rX   Úhuggingface_hub.dataclassesr   Úbackbone_utilsr   Úconfiguration_utilsr   Úutilsr   r   Ú__all__r   r.   r$   Ú<module>rd      sP   ðñ  å .å 1Ý 3Ý #ñ Ð4Ô5Øô^Ð%Ð'7ó ^ó ó 6ð^ðB ˆ/‚r.   