o ۷imP@svddlmZmZddlZddlmZddlmZmZddl 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(e)e*Z+Gddde Z,Gddde$Z-Gddde"Z.GdddeZ/Gddde Z0Gddde#Z1Gddde(Z2Gdd d 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) Gemma2ModelcseZdZdZdZdgZddddddddZdgdgfd d gd gfd gd gfd Z                   d fdd ZZ 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 s|_| _|_|_|_|_|_|dur|}|_|_| _ | _ | _ |_ |_ |_|_|_|_||_ttjd| |||d||dd_jdurstdd_fddtjD_tjjdS)N) pad_token_id bos_token_id eos_token_idtie_word_embeddingssliding_window_patterncs&g|]}t|djrdndqS)sliding_attentionfull_attention)bool_sliding_window_pattern).0iselfa/home/ubuntu/vllm_env/lib/python3.10/site-packages/transformers/models/cohere2/modular_cohere2.py sz*Cohere2Config.__init__..rF) 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_dimr super__init__getrAgetattrranger)rErIrKrMrLrNrOrPrQrJrRrSrTr7r8r9r:rUrVrWrXrYrZkwargs __class__rDrGr]sJ   zCohere2Config.__init__)r)r*r+r,r-r.Nr/r*r0r1Trr2r3Tr4NFr5r6N) __name__ __module__ __qualname____doc__ model_typekeys_to_ignore_at_inferencebase_model_tp_planbase_model_pp_planr] __classcell__rFrFrbrGr/sNp  rc@ eZdZdS)Cohere2RotaryEmbeddingNrdrerfrFrFrFrGrnrnc@rm)Cohere2LayerNormNrorFrFrFrGrqrprqc@seZdZdZddedeefddZeddd d   dd e j d e e j e j fd ee j dee dee j deede e j ee j ee e j ffddZdS)Cohere2Attentionz=Multi-headed attention from 'Attention Is All You Need' paperNconfig layer_idxcCstj|||_||_t|d|j|j|_|j|j |_ |jd|_ |j |_ d|_ |j|dkr5|jnd|_tj|j|j|j|jd|_tj|j|j |j|jd|_tj|j|j |j|jd|_tj|j|j|j|jd|_dS)Nr[gTr>)bias)nnModuler]rsrtr_rKrOr[rPnum_key_value_groupsscalingrX is_causalrZrYLinearrWq_projk_projv_projo_projrErsrtrFrFrGr]s*   zCohere2Attention.__init__past_key_valuer4.58new_nameversionr$position_embeddingsr%cache_positionrareturncKs2|jdd}g|d|jR}|||dd} |||dd} |||dd} |\} } |jdurGt| | | | \} } |dur\| | |d}| | | |j |\} } t }|j j dkrjt|j j }||| | | |f|jsvdn|j|j|jd|\}}|jg|dR}||}||fS)Nr=r)sincosreagerr5)dropoutryrY)shaper[r|view transposer}r~rYrupdatertrrs_attn_implementationrtrainingrXryreshape contiguousr)rEr$rr%rrra input_shape hidden_shape query_states key_states value_statesrr cache_kwargsattention_interface attn_output attn_weightsrFrFrGforward s<       zCohere2Attention.forwardN)NN)rdrerfrgrrintr]rtorchTensortupler LongTensorrr rrFrFrFrGrrs* rrcseZdZdedeffdd Zedddd dd ejd e ejejfd e ejde e de e de ej deede eje e ejejfffddZZS)Cohere2DecoderLayerrsrtcst|||j||_dSr)r\r]rZattention_typerrbrFrGr]OszCohere2DecoderLayer.__init__rrrrNFr$rr%rTrrarc KsJ|}||}|jd||||||d|\} } ||} || | }|S)N)r$rr%rrTrrF)input_layernorm self_attnmlp) rEr$rr%rrTrraresidualhidden_states_attention_hidden_states_mlprFrFrGrSs  zCohere2DecoderLayer.forward)NNFN)rdrerfrrr]rrrrrrr@rrr FloatTensorrrlrFrFrbrGrNs0  rc@seZdZUeed<dS)Cohere2PreTrainedModelrsN)rdrerfr__annotations__rFrFrFrGros rcseZdZdeffdd Z       ddeejdeejdeejdee d eej d ee d eejd e e d efddZZS) Cohere2Modelrscs.t|t|j|jd|_t|d|_dS)N)rKepsrs)r\r]rqrKrSr(rn rotary_emb)rErsrbrFrGr]ts zCohere2Model.__init__Nr"r% position_idsrr#rTrrarc Ks&|du|duAr td|dur||}|r$|dur$|js$t|jd}|dur@|dur0|nd} tj| | |jd|j d}|durI| d}t |} t si|j|||||d} t d i| td i| d} |} || |} |jD]}|| f| | |j|||d|} qt|| } t| |d S) Nz:You must specify exactly one of input_ids or inputs_embedsrrr=)device)rs input_embedsr%rrr)r?r>)rr%rrTr)last_hidden_staterrF) ValueErrorr&rrrsget_seq_lengthrarangerr unsqueeze isinstancedictr r rr'rr(r )rEr"r%rrr#rTrrapast_seen_tokenscausal_mask_mapping mask_kwargsr$r decoder_layerrFrFrGrysV        zCohere2Model.forward)NNNNNNN)rdrerfrr]rrrrrrr@rrr rrlrFrFrbrGrss8  rc@rm)Cohere2ForCausalLMNrorFrFrFrGrrpr)rrrr)5typingrrrtorch.nnrv cache_utilsrrconfiguration_utilsrr masking_utilsr r modeling_flash_attention_utilsr modeling_outputsr modeling_rope_utilsr modeling_utilsrprocessing_utilsrutilsrrutils.deprecationrcohere.modeling_cohererrrrrrrrgemma2.modeling_gemma2r get_loggerrdloggerrrnrqrrrrrr__all__rFrFrFrGs4       ( OI!E