hQddlZddlmZmZddlZddlmZddlmZm Z ddl m Z m Z ddl mZmZddlmZddlmZdd lmZdd lmZdd lmZdd lmZmZdd lmZddlm Z m!Z!m"Z"m#Z#m$Z$m%Z%m&Z&m'Z'ddl(m)Z)ejTe+Z,Gdde Z-Gdde%Z.Gdde#Z/Gdde Z0Gdde!Z1Gdde$Z2Gdde)Z3Gdd e"Z4gd!Z5y)"N)CallableOptional)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 Ze d Z e jdZ xZ 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_tokenslayersnormc ||_| |_||_||_||_||_||_||}||_||_| |_ | |_ | |_ ||_ ||_ ||_||_||_||_||z|_t'|t)|Td| |||d||j-dd|_|j"Wt1|dd|_t3|j Dcgc]!}t5|dz|j.zrdnd#c}|_t7|j"ycc}w)N) pad_token_id bos_token_id eos_token_idtie_word_embeddingssliding_window_patternsliding_attentionfull_attention) 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__get_sliding_window_patterngetattrrangeboolr )selfr6r8r:r9r;r<r=r>r7r?r@rAr,r-r.r/rBrCrDrErFrGkwargsi __class__s i/var/www/html/eduruby.in/venv/lib/python3.12/site-packages/transformers/models/cohere2/modular_cohere2.pyrJzCohere2Config.__init__sm4%'>$&&!2!2#6   &"5 #6 $!2,"$(,!2,&#'::  t$  %%% 3    (.zz2JA'N$    #+249QST+UD (t556 (,QUd6R6R,R'S#Yii D  d../  s9&D=cNtjdt|jS)NzTThe `sliding_window_pattern` attribute is deprecated and will be removed in v4.55.0.)warningswarn FutureWarningrL)rPs rTr0z$Cohere2Config.sliding_window_patterns" b  +++c||_yN)rL)rPvalues rTr0z$Cohere2Config.sliding_window_patterns ',$rY)i iXg?(@Nsilur]g{Gz?gh㈵>TriTg@NFiN) __name__ __module__ __qualname____doc__ model_typekeys_to_ignore_at_inferencebase_model_tp_planbase_model_pp_planrJpropertyr0setter __classcell__rSs@rTrr0sn`J#4"5%.%.%.%."+ )"+&(9:#%568IJ!"_$56  $ /I0V,,""-#-rYrc eZdZy)Cohere2RotaryEmbeddingNrcrdrer5rYrTrprp rYrpc eZdZy)Cohere2LayerNormNrqr5rYrTrtrtrrrYrtcNeZdZdZddedeefdZeddd  dd e jd e e je jfd ee jdee d ee jdeede e jee jee e jffdZy)Cohere2Attentionz=Multi-headed attention from 'Attention Is All You Need' paperNconfig layer_idxctjj|||_||_t |d|j |jz|_|j|jz|_ |jdz|_ |j|_ d|_ |j|dk(r |jnd|_tj |j |j|jz|j"|_tj |j |j|jz|j"|_tj |j |j|jz|j"|_tj |j|jz|j |j"|_y)NrHgTr3)bias)nnModulerJrwrxrMr8r<rHr=num_key_value_groupsscalingrE is_causalrGrFLinearrDq_projk_projv_projo_proj)rPrwrxs rTrJzCohere2Attention.__init__sw 4  " F4F4F&JdJd4de $*$>$>&B\B\$\!}}d* !'!9!97=7I7I)7TXk7kf33quii    : :T]] JQWQfQf  ii    : :T]] JQWQfQf  ii    : :T]] JQWQfQf  ii  & & 68J8JQWQfQf  rYpast_key_valuer!4.58new_nameversionr&position_embeddingsr'cache_positionrQreturnc b|jdd}g|d|j}|j|j|j dd} |j |j|j dd} |j |j|j dd} |\} } |jt| | | | \} } |'| | |d}|j| | |j|\} } t}|jjdk7rt|jj}||| | | |f|jsdn |j |j"|jd|\}}|j$g|dj'}|j)|}||fS)Nr2r)sincosreagerrb)dropoutr~rF)shaperHrview transposerrrFrupdaterxrrw_attn_implementationrtrainingrEr~reshape contiguousr)rPr&rr'r!rrQ input_shape hidden_shape query_states key_states value_statesrr cache_kwargsattention_interface attn_output attn_weightss rTforwardzCohere2Attention.forward-s$))#2. 88b8$--8 {{=166|DNNqRST [[/44\BLLQPQR {{=166|DNNqRST &S    *';L*VY[^'_ $L*  &#&snUL'6'='=j,X\XfXfht'u $J (? ;; + +w 6"9$++:Z:Z"[ $7      % $}}C$2H2HLL.. %  % ! \*k));;;;FFH kk+. L((rYr[)NN)rcrdrerfrrintrJrtorchTensortupler LongTensorrr rr5rYrTrvrvsG } # 0%0A6R ,059 *)||*)#5<<#=>*)!. *) "% *) !!1!12 *)-.*) u||Xell3XeELL>Q5RR S*)S*)rYrvcXeZdZdedeffd Zeddd ddejd e ejejfd e ejde e d e e d e ejd eede ej e e ej ej fffdZxZS)Cohere2DecoderLayerrwrxcNt||||j||_yr[)rIrJrGattention_type)rPrwrxrSs rTrJzCohere2DecoderLayer.__init__\s% +$00;rYrr!rrr&rr'rArrQrc |}|j|}|jd||||||d|\} } |j|} || z| z}|S)N)r&rr'r!rArr5)input_layernorm self_attnmlp) rPr&rr'r!rArrQresidualhidden_states_attention_hidden_states_mlps rTrzCohere2DecoderLayer.forward`sx!,,]; %3T^^& ' 3)+) & & "!HH]3 #::=NN rY)NNFN)rcrdrerrrJrrrrrrrOrrr FloatTensorrrmrns@rTrr[s<}<<%0A6R 26+/$)59||#5<<#=>!.  "%  D> !!1!12-. u  (51B1BEDUDU1U+V"WW XSrYrceZdZUeed<y)Cohere2PreTrainedModelrwN)rcrdrer__annotations__r5rYrTrr|s rYrceZdZdeffd Z d deejdeejdeejdee deejdee d eejd e e d efd ZxZS) Cohere2Modelrwct||t|j|j|_t ||_y)N)r8epsrw)rIrJrtr8r@r*rp rotary_emb)rPrwrSs rTrJzCohere2Model.__init__s6  $&2D2D6K`K`a 0?rYr$r' position_idsr!r%rArrQrc |du|duz r td||j|}|r$|"|jst|j}|F||j nd} t j| | |jdz|j}||jd}t|x} ts*|j|||||d} td i| td i| d} |} |j| |} |j D]}|| f| | |j"|||d|} |j%| } t'| | S) Nz:You must specify exactly one of input_ids or inputs_embedsrrr2)device)rw input_embedsr'rr!r)r4r3)rr'r!rAr)last_hidden_stater!r5) ValueErrorr(rrrwget_seq_lengthrarangerr unsqueeze isinstancedictr r rr)rr*r )rPr$r'rr!r%rArrQpast_seen_tokenscausal_mask_mapping mask_kwargsr&r decoder_layers rTrzCohere2Model.forwards  -t"; <YZ Z   --i8M 0*$++>O  !CRC^==?de "\\ "2]5H5H5K"KTaThThN  )33A6L?-F++ -"0"0#2 , K#5"C{"C%F%U%U#  & "oom\J![[ M)$72=3O3OP /#- M  -0 &++  rY)NNNNNNN)rcrdrerrJrrrrrrrOrrr rrmrns@rTrrs@}@151537+/59$(59< E,,-< !.< u//0 < "% <   1 12 < D>< !!1!12< +,<  !< rYrc eZdZy)Cohere2ForCausalLMNrqr5rYrTrrrrrYr)rrrr)6rVtypingrrrtorch.nnr{ cache_utilsrrconfiguration_utilsrr masking_utilsr r modeling_flash_attention_utilsr modeling_outputsr modeling_rope_utilsrmodeling_utilsrprocessing_utilsrutilsrrutils.deprecationrcohere.modeling_cohererrrrrrrrgemma2.modeling_gemma2r get_loggerrcloggerrrprtrvrrrr__all__r5rYrTrs % .JRB795&00   1   H %W-$W-t 2   F)F)R,B2B ;B J *  \rY