+
    œ°³j—  ã                   óz   € R t ^ RIHt ^RIHt ^RIHt ^RIHt ]]! RRR7       ! R R	]4      4       4       t	R	.t
R
# )zBamba model configuration)Ústrict)ÚPreTrainedConfig)ÚRopeParameters)Úauto_docstringz´
    The BambaModel is a hybrid [mamba2](https://github.com/state-spaces/mamba) architecture with SwiGLU.
    The checkpoints are  jointly trained by IBM, Princeton, and UIUC.
    zibm-fms/Bamba-9.8b-2.2T-hf)Úcustom_introÚ
checkpointc                   óú   a a€ ] tR t^t oRtRtRR/tR.tRtRt	Rt
R	t^ t^ t^tR
tRtRtRt^t^ t^t^tRtRtRt^€tRt^tRt^t^tRt Rt!Rt"R]#! R4      3t$Rt%Rt&Rt'Rt(V 3R lt)]*R 4       t+R t,V3R lt-Rt.Vt/V ;t0# )ÚBambaConfiga"  
num_logits_to_keep (`int` or `None`, *optional*, defaults to 1):
    Number of prompt logits to calculate during generation. If `None`, all logits will be calculated. If an
    integer value, only last `num_logits_to_keep` logits will be calculated. Default is 1 because only the
    logits of the last prompt token are needed for generation. For long sequences, the logits for the entire
    sequence may use a lot of memory so, setting `num_logits_to_keep=1` will reduce memory footprint
    significantly.
attn_layer_indices (`list`, *optional*):
    Specifies the layer indices that will have full attention. Must contain values at most num_hidden_layers.
z_loss_coefficient (`float`, *optional*, defaults to 0.0):
    Coefficient for auxiliary z-loss used to control logit growth during training
ÚbambaÚlayer_typesÚlayers_block_typeÚpast_key_valuesi ô Fi   i 8  Úsilug{®Gáz”?gñhãˆµøä>Ti   g        NÚautoé   Úinfc                óL  <€ V P                   f   V P                  V n         V P                  R8X  d4   V P                  V P                  ,          V P
                  ,          V n        V P                  e   \        V P                  4      MR V n        RVR&   \        SV `$  ! R/ VB  R # )Nr   g      à?Úpartial_rotary_factor© )
Únum_key_value_headsÚnum_attention_headsÚmamba_d_headÚmamba_expandÚhidden_sizeÚmamba_n_headsÚtime_step_limitÚtupleÚsuperÚ__post_init__)ÚselfÚkwargsÚ	__class__s   &,€Ún/home/luca/screen-checking/.venv/lib/python3.14/site-packages/transformers/models/bamba/configuration_bamba.pyr   ÚBambaConfig.__post_init__R   s�   ø€ à×#Ñ#Ò+Ø'+×'?Ñ'?ˆDÔ$ð ×Ñ Ô&Ø $× 1Ñ 1°D×4DÑ4DÕ DÈ×HZÑHZÕ ZˆDÔà>B×>RÑ>RÒ>^œu T×%9Ñ%9Ô:ÐdhˆÔØ*-ˆÐ&Ñ'ä‰ÒÑ' Ô'ó    c                óž   € \        V P                  4       Uu. uF)  pV P                  '       d   WP                  9   d   R MRNK+  	  up# u upi )Úfull_attentionÚlinear_attention)ÚrangeÚnum_hidden_layersÚattn_layer_indices)r   Úis   & r"   r   ÚBambaConfig.layers_block_type`   sQ   € ô ˜4×1Ñ1Ô2ó
á2�ð "&×!8×!8Ð!8¸Q×BYÑBYÔ=YÑÐ`rÒrÙ2ñ
ð 	
ùò 
s
   ˜A
¯A
c                óÞ   € V P                   V P                  ,          pWP                  ,          ^ 8w  d   \        R4      hV P                  V P                  ,          V8w  d   \        R4      hR# )zOPart of `@strict`-powered validation. Validates the architecture of the config.z4mamba_n_heads must divide mamba_expand * hidden_sizezPThe dimensions for the Mamba head state do not match the model intermediate_sizeN)r   r   r   Ú
ValueErrorr   )r   Úmamba_intermediates   & r"   Úvalidate_architectureÚ!BambaConfig.validate_architectureg   sb   € à!×.Ñ.°×1AÑ1AÕAÐØ× 2Ñ 2Õ2°aÔ7ÜÐSÓTÐTà×Ñ˜t×1Ñ1Õ1Ð5GÔGÜÐoÓpÐpñ Hr$   c                ó:  <€ V ^8„  d   Qh/ S[ ;R&   S[;R&   S[ ;R&   S[ ;R&   S[ ;R&   S[ ;R&   S[ R,          ;R&   S[;R	&   S[;R
&   S[;R&   S[;R&   S[ R,          ;R&   S[ R,          ;R&   S[ R,          ;R&   S[ S[S[ ,          ,          R,          ;R&   S[ ;R&   S[S[ ,          R,          ;R&   S[S[ ,          R,          ;R&   S[ R,          ;R&   S[S[ ,          R,          ;R&   S[ R,          ;R&   S[ R,          ;R&   S[ R,          ;R&   S[ R,          ;R&   S[ R,          ;R&   S[R,          ;R&   S[R,          ;R&   S[S[,          S[S[S[3,          ,          R,          ;R&   S[R,          ;R&   S[S[,          R,          ;R&   S[;R &   S[;R!&   # )"é   Ú
vocab_sizeÚtie_word_embeddingsr   Úintermediate_sizer)   r   Nr   Ú
hidden_actÚinitializer_rangeÚrms_norm_epsÚ	use_cacheÚnum_logits_to_keepÚpad_token_idÚbos_token_idÚeos_token_idÚmax_position_embeddingsÚattention_dropoutr*   r   r   Úmamba_n_groupsÚmamba_d_stateÚmamba_d_convr   Úmamba_chunk_sizeÚmamba_conv_biasÚmamba_proj_biasr   Úz_loss_coefficientÚrope_parametersÚattention_biasÚmlp_bias)ÚintÚboolÚstrÚfloatÚlistr   r   Údict)ÚformatÚ__classdict__s   "€r"   Ú__annotate__ÚBambaConfig.__annotate__   sV  ø‡ ‚ ñ4 Ññ5 ñ6 Ñ%ñ7 ñ8 Ññ9 ñ: Ñ"ñ; ñ< Ññ= ñ> Ñ!ñ? ñ@ ˜t�Ñ'ñA ñB ÑñC ñD Ñ#ñE ñF ÑñG ñH ÑñI ñJ ˜d�
Ñ&ñK ñL ˜•*Ñ ñM ñN ˜•*Ñ ñO ñP ™™S�	•/ DÕ(Ñ,ñQ ñR !Ñ)ñS ñT ™s•{ TÕ)Ñ/ñU ñV ™S�	 DÕ(Ñ/ñW ñX ˜•:Ñ#ñY ñZ ™•)˜dÕ"Ñ+ñ[ ñ\ ˜$•JÑ"ñ] ñ^ ˜•:Ñ#ñ_ ñ` ˜•*Ñ ña ñb ˜•*Ñ ñc ñd ˜D•jÑ&ñe ñf ˜D•[Ñ'ñg ñh ˜D•[Ñ(ñi ñj ™%•[¡5©±¨Õ#6Õ6¸Õ=ÑSñk ñl  �Ñ*ñm ñn $¡dÕ*¨TÕ1Ñ8ño ñp Ñ ñq ñr Ñòs r$   )r   r   r   )1Ú__name__Ú
__module__Ú__qualname__Ú__firstlineno__Ú__doc__Ú
model_typeÚattribute_mapÚkeys_to_ignore_at_inferencer4   r5   r   r6   r)   r   r   r7   r8   r9   r:   r;   r<   r=   r>   r?   r@   r*   r   r   rA   rB   rC   r   rD   rE   rF   rN   r   rG   rH   rI   rJ   r   Úpropertyr   r0   Ú__annotate_func__Ú__static_attributes__Ú__classdictcell__Ú__classcell__)r!   rR   s   @@r"   r	   r	      s  ù‡ € ñð €JØ"Ð$7Ð8€MØ#4Ð"5Ðà€JØ %ÐØ€KØ"ÐØÐØ!ÐØ&'ÐØ€JØ#ÐØ€LØ€IØ%&ÐØ €LØ €LØ+,€LØ#)ÐØ,/ÐØ+/ÐØ #€MØ%+€LØ!"€NØ #€MØ €LØ €LØ#&ÐØ#'€OØ#(€OØADÁeÈEÃlÐ@S€OØ'*ÐØ48€OØ €NØ€Hõ(ð ñ
ó ð
òq÷a … r$   r	   N)rY   Úhuggingface_hub.dataclassesr   Úconfiguration_utilsr   Úmodeling_rope_utilsr   Úutilsr   r	   Ú__all__r   r$   r"   Ú<module>rg      sY   ðñ  å .å 3Ý 1Ý #ð Ùðð ,ôôOqÐ"ó Oqóó ðOqðd ˆ/‚r$   