`imPddlmZmZddlZddlmZddlmZmZddl m Z m Z ddl m Z mZddlmZddlmZdd lmZdd lmZdd lmZdd lmZmZdd lmZddlmZm Z m!Z!m"Z"m#Z#m$Z$m%Z%m&Z&ddl'm(Z(ej)e*Z+Gdde Z,Gdde$Z-Gdde"Z.GddeZ/Gdde Z0Gdde#Z1Gdde(Z2Gdd e!Z3gd!Z4dS)")CallableOptionalN)Cache DynamicCache)PretrainedConfiglayer_type_validation)create_causal_mask!create_sliding_window_causal_mask)FlashAttentionKwargs)BaseModelOutputWithPast)rope_config_validation)ALL_ATTENTION_FUNCTIONS)Unpack)TransformersKwargslogging)deprecate_kwarg)CohereAttentionCohereDecoderLayerCohereForCausalLMCohereLayerNormCoherePreTrainedModelCohereRotaryEmbeddingapply_rotary_pos_embeager_attention_forward) Gemma2ModelceZdZdZdZdgZddddddddZdgdgfd d gd gfd gd gfd Z dfd ZxZ S) Cohere2Configa2 This is the configuration class to store the configuration of a [`CohereModel`]. It is used to instantiate an Cohere model according to the specified arguments, defining the model architecture. Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the documentation from [`PretrainedConfig`] for more information. Instantiating a configuration with the defaults will yield a similar configuration to that of the [CohereForAI/c4ai-command-r-v01](https://huggingface.co/CohereForAI/c4ai-command-r-v01) model. Args: vocab_size (`int`, *optional*, defaults to 256000): Vocabulary size of the Cohere model. Defines the number of different tokens that can be represented by the `inputs_ids` passed when calling [`CohereModel`] hidden_size (`int`, *optional*, defaults to 8192): Dimension of the hidden representations. intermediate_size (`int`, *optional*, defaults to 22528): Dimension of the MLP representations. logit_scale (`float`, *optional*, defaults to 0.0625): The scaling factor for the output logits. num_hidden_layers (`int`, *optional*, defaults to 40): Number of hidden layers in the Transformer decoder. num_attention_heads (`int`, *optional*, defaults to 64): Number of attention heads for each attention layer in the Transformer decoder. num_key_value_heads (`int`, *optional*): This is the number of key_value heads that should be used to implement Grouped Query Attention. If `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if `num_key_value_heads=1` the model will use Multi Query Attention (MQA) otherwise GQA is used. When converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed by meanpooling all the original heads within that group. For more details, check out [this paper](https://huggingface.co/papers/2305.13245). If it is not specified, will default to `num_attention_heads`. hidden_act (`str` or `function`, *optional*, defaults to `"silu"`): The non-linear activation function (function or string) in the decoder. max_position_embeddings (`int`, *optional*, defaults to 8192): The maximum sequence length that this model might ever be used with. initializer_range (`float`, *optional*, defaults to 0.02): The standard deviation of the truncated_normal_initializer for initializing all weight matrices. layer_norm_eps (`float`, *optional*, defaults to 1e-05): The epsilon used by the layer normalization. use_cache (`bool`, *optional*, defaults to `True`): Whether or not the model should return the last key/values attentions (not used by all models). Only relevant if `config.is_decoder=True`. pad_token_id (`int`, *optional*, defaults to 0): Padding token id. bos_token_id (`int`, *optional*, defaults to 5): Beginning of stream token id. eos_token_id (`int`, *optional*, defaults to 255001): End of stream token id. tie_word_embeddings (`bool`, *optional*, defaults to `True`): Whether to tie weight embeddings rope_theta (`float`, *optional*, defaults to 10000.0): The base period of the RoPE embeddings. rope_scaling (`Dict`, *optional*): Dictionary containing the scaling configuration for the RoPE embeddings. NOTE: if you apply new rope type and you expect the model to work on longer `max_position_embeddings`, we recommend you to update this value accordingly. Expected contents: `rope_type` (`str`): The sub-variant of RoPE to use. Can be one of ['default', 'linear', 'dynamic', 'yarn', 'longrope', 'llama3'], with 'default' being the original RoPE implementation. `factor` (`float`, *optional*): Used with all rope types except 'default'. The scaling factor to apply to the RoPE embeddings. In most scaling types, a `factor` of x will enable the model to handle sequences of length x * original maximum pre-trained length. `original_max_position_embeddings` (`int`, *optional*): Used with 'dynamic', 'longrope' and 'llama3'. The original max position embeddings used during pretraining. `attention_factor` (`float`, *optional*): Used with 'yarn' and 'longrope'. The scaling factor to be applied on the attention computation. If unspecified, it defaults to value recommended by the implementation, using the `factor` field to infer the suggested value. `beta_fast` (`float`, *optional*): Only used with 'yarn'. Parameter to set the boundary for extrapolation (only) in the linear ramp function. If unspecified, it defaults to 32. `beta_slow` (`float`, *optional*): Only used with 'yarn'. Parameter to set the boundary for interpolation (only) in the linear ramp function. If unspecified, it defaults to 1. `short_factor` (`list[float]`, *optional*): Only used with 'longrope'. The scaling factor to be applied to short contexts (< `original_max_position_embeddings`). Must be a list of numbers with the same length as the hidden size divided by the number of attention heads divided by 2 `long_factor` (`list[float]`, *optional*): Only used with 'longrope'. The scaling factor to be applied to long contexts (< `original_max_position_embeddings`). Must be a list of numbers with the same length as the hidden size divided by the number of attention heads divided by 2 `low_freq_factor` (`float`, *optional*): Only used with 'llama3'. Scaling factor applied to low frequency components of the RoPE `high_freq_factor` (`float`, *optional*): Only used with 'llama3'. Scaling factor applied to high frequency components of the RoPE attention_bias (`bool`, defaults to `False`, *optional*, defaults to `False`): Whether to use a bias in the query, key, value and output projection layers during self-attention. attention_dropout (`float`, *optional*, defaults to 0.0): The dropout ratio for the attention probabilities. sliding_window (`int`, *optional*, defaults to 4096): Size of the sliding window attention context. layer_types (`list`, *optional*): Attention pattern for each layer. ```python >>> from transformers import Cohere2Model, Cohere2Config >>> # Initializing a Cohere Nextmodel configuration >>> configuration = Cohere2Config() >>> # Initializing a model from the Cohere2 configuration >>> model = Cohere2Model(configuration) # doctest: +SKIP >>> # Accessing the model configuration >>> configuration = model.config # doctest: +SKIP ``` cohere2past_key_valuescolwiserowwise)zlayers.*.self_attn.q_projzlayers.*.self_attn.k_projzlayers.*.self_attn.v_projzlayers.*.self_attn.o_projzlayers.*.mlp.gate_projzlayers.*.mlp.up_projzlayers.*.mlp.down_proj input_ids inputs_embeds hidden_statesattention_mask) embed_tokenslayersnorm X?(@Nsilu{Gz?h㈵>Tr@Fc j|_| _|_|_|_|_|_||}|_|_| _ | _ | _ |_ |_ |_|_|_|_||z_t't)jd| |||d||dd_j;t1dd_fdt3jD_t5jjdS)N) pad_token_id bos_token_id eos_token_idtie_word_embeddingssliding_window_patterncLg|] }t|dzjzrdnd!S)sliding_attentionfull_attention)bool_sliding_window_pattern).0iselfs /home/jaya/work/projects/VOICE-AGENT/VIET/agent-env/lib/python3.11/site-packages/transformers/models/cohere2/modular_cohere2.py z*Cohere2Config.__init__..sI   (,QUd6R,R'S'Si##Yi   ) vocab_sizemax_position_embeddings hidden_size logit_scaleintermediate_sizenum_hidden_layersnum_attention_headsnum_key_value_heads hidden_actinitializer_rangelayer_norm_eps use_cache rope_theta rope_scalingattention_biasattention_dropoutsliding_window layer_typeshead_dimrsuper__init__getrEgetattrranger )rHrMrOrQrPrRrSrTrUrNrVrWrXr:r;r<r=rYrZr[r\r]r^kwargs __class__s` rIrazCohere2Config.__init__s4%'>$&&!2!2#6   &"5 #6 $!2,"$(,!2,&#'::  t$$$ %%% 3       (.zz2JA'N'N$   #+249QST+U+UD (    t566   D  d.0FGGGGGrK)r+r,r-r.r/r0Nr1r,r2r3Trr4r5Tr6NFr7r8N) __name__ __module__ __qualname____doc__ model_typekeys_to_ignore_at_inferencebase_model_tp_planbase_model_pp_planra __classcell__rfs@rIrr/snn`J#4"5%.%.%.%."+ )"+&(9:#%568IJ!"_$56  $ /IHIHIHIHIHIHIHIHIHIHrKrceZdZdS)Cohere2RotaryEmbeddingNrgrhrirLrKrIrrrrDrKrrceZdZdS)Cohere2LayerNormNrsrLrKrIrvrvrtrKrvcHeZdZdZddedeefdZeddd  dd e j d e e j e j fd ee j dee d ee j deede e j ee j ee e j ffdZdS)Cohere2Attentionz=Multi-headed attention from 'Attention Is All You Need' paperNconfig layer_idxctj|||_||_t |d|j|jz|_|j|j z|_ |jdz|_ |j |_ d|_ |j|dkr|jnd|_tj|j|j|jz|j|_tj|j|j |jz|j|_tj|j|j |jz|j|_tj|j|jz|j|j|_dS)Nr_gTrB)bias)nnModuleraryrzrcrOrSr_rTnum_key_value_groupsscalingr\ is_causalr^r]Linearr[q_projk_projv_projo_proj)rHryrzs rIrazCohere2Attention.__init__se 4    " F4F&Jd4dee $*$>&B\$\!}d* !'!97=7I)7TXk7k7kf33qui   :T] JQWQf    i   :T] JQWQf    i   :T] JQWQf    i  & 68JQWQf    rKpast_key_valuer!4.58new_nameversionr&position_embeddingsr'cache_positionrereturnc ^|jdd}g|d|jR}|||dd} |||dd} |||dd} |\} } |jt| | | | \} } |&| | |d}| | | |j |\} } t}|j j dkrt|j j }||| | | |f|jsdn|j|j|jd|\}}|jg|dR}||}||fS)NrAr)sincosreagerr7)dropoutrr])shaper_rview transposerrr]rupdaterzrry_attn_implementationrtrainingr\rreshape contiguousr)rHr&rr'r!rre input_shape hidden_shape query_states key_states value_statesrr cache_kwargsattention_interface attn_output attn_weightss rIforwardzCohere2Attention.forward s$)#2#. 88b8$-88 {{=1166|DDNNqRSTT [[//44\BBLLQPQRR {{=1166|DDNNqRSTT &S   *';L*VY[^'_'_ $L*  &#&snUUL'6'='=j,X\Xfht'u'u $J (? ; +w 6 6"9$+:Z"[ $7$7      % $}HCC$2HL. % %  % % ! \*k);;;;;;FFHH kk+.. L((rKN)NN)rgrhrirjrrintrartorchTensortupler LongTensorrr rrLrKrIrxrxsGG  } #    0_%0A6RRR ,059 *)*)|*)#5<#=>*)!. *) "% *) !!12 *)-.*) u|Xel3XeEL>Q5RR S*)*)*)SR*)*)*)rKrxcJeZdZdedeffd Zeddd dd ejd e ejejfd e ejde e d e e de ej deede eje e ejejfffdZxZS)Cohere2DecoderLayerryrzcpt|||j||_dSr)r`rar^attention_type)rHryrzrfs rIrazCohere2DecoderLayer.__init__Os3 +++$0;rKrr!rrNFr&rr'rXrrerc |}||}|jd||||||d|\} } ||} || z| z}|S)N)r&rr'r!rXrrL)input_layernorm self_attnmlp) rHr&rr'r!rXrreresidualhidden_states_attention_hidden_states_mlps rIrzCohere2DecoderLayer.forwardSs!,,];; %3T^& ' 3)+) & & & & "!HH]33 #::=NN rK)NNFN)rgrhrirrrarrrrrrrDrrr FloatTensorrrorps@rIrrNs0<}<<<<<<<_%0A6RRR 26+/$)59|#5<#=>!.  "%  D> !!12-. u (51BEDU1U+V"WW XSRrKrceZdZUeed<dS)Cohere2PreTrainedModelryN)rgrhrir__annotations__rLrKrIrros rKrceZdZdeffd Z ddeejdeejdeejdee deej d ee d eejd e e d efd ZxZS) Cohere2Modelryct|t|j|j|_t ||_dS)N)rOepsry)r`rarvrOrWr*rr rotary_emb)rHryrfs rIrazCohere2Model.__init__tsK    $&2D6K`aaa 0???rKNr$r' position_idsr!r%rXrrerc |du|duz rtd|||}|r||jst|j}|B||nd} t j| | |jdz|j }|| d}t|x} ts'|j|||||d} td i| td i| d} |} || |} |jD]}|| f| | |j|||d|} || } t'| | S) Nz:You must specify exactly one of input_ids or inputs_embedsrrrA)device)ry input_embedsr'rr!r)rCrB)rr'r!rXr)last_hidden_stater!rL) ValueErrorr(rrryget_seq_lengthrarangerr unsqueeze isinstancedictr r rr)rr*r )rHr$r'rr!r%rXrrepast_seen_tokenscausal_mask_mapping mask_kwargsr&r decoder_layers rIrzCohere2Model.forwardys  -t"; < [YZZ Z   --i88M  ?00*$+>>>O  !CRC^==???de "\ "2]5H5K"KTaThN  )33A66L?-FF + -"0"0#2 , K#5"C"C{"C"C%F%U%U%U%U##  & "oom\JJ![  M)M$72=3OP /#- MM -00 &++    rK)NNNNNNN)rgrhrirrarrrrrrrDrrr rrorps@rIrrss@}@@@@@@151537+/59$(59< < E,-< !.< u/0 < "% <   12 < D>< !!12< +,<  !< < < < < < < < rKrceZdZdS)Cohere2ForCausalLMNrsrLrKrIrrrtrKr)rrrr)5typingrrrtorch.nnr} cache_utilsrrconfiguration_utilsrr masking_utilsr r modeling_flash_attention_utilsr modeling_outputsr modeling_rope_utilsrmodeling_utilsrprocessing_utilsrutilsrrutils.deprecationrcohere.modeling_cohererrrrrrrrgemma2.modeling_gemma2r get_loggerrgloggerrrrrvrxrrrr__all__rLrKrIrs &%%%%%%% ........JJJJJJJJRRRRRRRRBBBBBB777777999999555555&&&&&&00000000000000                    100000  H % %KHKHKHKHKH$KHKHKH\     2           F)F)F)F)F)F)F)F)R,B2B B B B B ;B B B J     *    \ [ [rK