o eid0@sddlZddlmZddlmZmZddlmZddlmZm Z ddl m Z ddl m Z dd lmZdd lmZmZdd lmZdd lmZd dlmZd dlmZmZmZmZmZmZm Z m!Z!m"Z"m#Z#m$Z$GdddeeZ%Gddde Z&GdddeZ'Gddde#Z(GdddeZ)Gddde"Z*Gddde$Z+Gddde!Z,Gd d!d!eZ-Gd"d#d#eZ.Gd$d%d%eZ/Gd&d'd'eZ0gd(Z1dS))N)nn)Cache DynamicCache)PreTrainedConfig)create_causal_mask!create_sliding_window_causal_mask)BaseModelOutputWithPast)RopeParameters)Unpack)TransformersKwargsauto_docstring)merge_with_config_defaults)capture_outputs) MistralConfig) Qwen2AttentionQwen2DecoderLayerQwen2ForCausalLMQwen2ForQuestionAnsweringQwen2ForSequenceClassificationQwen2ForTokenClassificationQwen2MLP Qwen2ModelQwen2PreTrainedModel Qwen2RMSNormQwen2RotaryEmbeddingc*@seZdZdZdZ               d(dedBdedBdedBdedBdedBdedBdedBdedBdedBdedBdedBdedBdedBdedBd edBd!edBd"e dBd#edBd$edBd%e edBf(d&d'Z dS))MinistralConfiga This is the configuration class to store the configuration of a [`MinistralModel`]. It is used to instantiate an Ministral model according to the specified arguments, defining the model architecture. Instantiating a configuration with the defaults will yield a similar configuration to that of the Ministral-8B-Instruct-2410. [mistralai/Ministral-8B-Instruct-2410](https://huggingface.co/mistralai/Ministral-8B-Instruct-2410) [mistralai/Ministral-8B-Instruct-2410](https://huggingface.co/mistralai/Ministral-8B-Instruct-2410) Configuration objects inherit from [`PreTrainedConfig`] and can be used to control the model outputs. Read the documentation from [`PreTrainedConfig`] for more information. Args: vocab_size (`int`, *optional*, defaults to 32000): Vocabulary size of the Ministral model. Defines the number of different tokens that can be represented by the `inputs_ids` passed when calling [`MinistralModel`] hidden_size (`int`, *optional*, defaults to 4096): Dimension of the hidden representations. intermediate_size (`int`, *optional*, defaults to 14336): Dimension of the MLP representations. num_hidden_layers (`int`, *optional*, defaults to 32): Number of hidden layers in the Transformer encoder. num_attention_heads (`int`, *optional*, defaults to 32): Number of attention heads for each attention layer in the Transformer encoder. num_key_value_heads (`int`, *optional*, defaults to 8): 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 `8`. head_dim (`int`, *optional*, defaults to `hidden_size // num_attention_heads`): The attention head dimension. 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 `4096*32`): The maximum sequence length that this model might ever be used with. Ministral's sliding window attention allows sequence of up to 4096*32 tokens. initializer_range (`float`, *optional*, defaults to 0.02): The standard deviation of the truncated_normal_initializer for initializing all weight matrices. rms_norm_eps (`float`, *optional*, defaults to 1e-06): The epsilon used by the rms normalization layers. 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*): The id of the padding token. bos_token_id (`int`, *optional*, defaults to 1): The id of the "beginning-of-sequence" token. eos_token_id (`int`, *optional*, defaults to 2): The id of the "end-of-sequence" token. tie_word_embeddings (`bool`, *optional*, defaults to `False`): Whether the model's input and output word embeddings should be tied. rope_parameters (`RopeParameters`, *optional*): Dictionary containing the configuration parameters for the RoPE embeddings. The dictionary should contain a value for `rope_theta` and optionally parameters used for scaling in case you want to use RoPE with longer `max_position_embeddings`. sliding_window (`int`, *optional*, defaults to 4096): Sliding window attention window size. If not specified, will default to `4096`. attention_dropout (`float`, *optional*, defaults to 0.0): The dropout ratio for the attention probabilities. layer_types (`list`, *optional*): Attention pattern for each layer. ```python >>> from transformers import MinistralModel, MinistralConfig >>> # Initializing a Ministral 8B style configuration >>> configuration = MinistralConfig() >>> # Initializing a model from the Ministral 8B style configuration >>> model = MinistralModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ``` ministral}8 Nsilu{Gz?ư>TrF vocab_size hidden_sizeintermediate_sizenum_hidden_layersnum_attention_headsnum_key_value_headshead_dim hidden_actmax_position_embeddingsinitializer_range rms_norm_eps use_cache pad_token_id bos_token_id eos_token_idtie_word_embeddingsrope_parameterssliding_windowattention_dropout layer_typescKs| |_||_||_||_||_| |_||_||_||_||_ ||_ ||_ |dur*|}||_ ||_ | |_| |_| |_||_||_|jdurQ|j durKdndg||_||_tj|fi|dS)Nsliding_attentionfull_attention)r6r7r8r9r*r2r+r,r-r.r;r0r/r1r3r4r5r<r=r:r__init__)selfr*r+r,r-r.r/r0r1r2r3r4r5r6r7r8r9r:r;r<r=kwargsrCm/home/ubuntu/transcripts/venv/lib/python3.10/site-packages/transformers/models/ministral/modular_ministral.pyr@ms8 zMinistralConfig.__init__)rr r!r"r"r#Nr$r%r&r'TNr(rFNr r)N) __name__ __module__ __qualname____doc__ model_typeintstrfloatboolr listr@rCrCrCrDrsM      rc@ eZdZdS) MinistralMLPNrErFrGrCrCrCrDrPrPcs"eZdZdeffdd ZZS)MinistralAttention layer_idxcsft||tj|j|j|jdd|_tj|j|j|jdd|_ tj|j|j|jdd|_ dS)NF)bias) superr@rLinearr+r.r0q_projr/k_projv_proj)rAconfigrT __class__rCrDr@s zMinistralAttention.__init__)rErFrGrJr@ __classcell__rCrCr\rDrSsrSc@rO)MinistralRMSNormNrQrCrCrCrDr_rRr_c@rO)MinistralDecoderLayerNrQrCrCrCrDr`rRr`c@rO)MinistralPreTrainedModelNrQrCrCrCrDrarRrac@rO)MinistralRotaryEmbeddingNrQrCrCrCrDrbrRrbcseZdZdeffdd Zeee       ddej dBdej dBdej dBde dBd ej dBd e dBd ej dBd eed efddZZS)MinistralModelr[cst||`dS)N)rVr@has_sliding_layers)rAr[r\rCrDr@s zMinistralModel.__init__N input_idsattention_mask position_idspast_key_values inputs_embedsr5cache_positionrBreturnc Ks:|du|duAr td|dur||}|r!|dur!t|jd}|dur=|dur-|nd} tj| | |jd|jd}|durF| d}t |} t sf|j|||||d} t d i| t d i| d} |} || |} |jd|jjD]}|| f| |j||||| d|} qw|| } t| |r|d Sdd S) Nz:You must specify exactly one of input_ids or inputs_embeds)r[rr()device)r[rirfrjrhrg)r?r>)rfrgrhr5rjposition_embeddings)last_hidden_staterhrC) ValueError embed_tokensrr[get_seq_lengthtorcharangeshaperl unsqueeze isinstancedictrr rotary_emblayersr-attention_typenormr )rArerfrgrhrir5rjrBpast_seen_tokenscausal_mask_mapping mask_kwargs hidden_statesrm decoder_layerrCrCrDforwards\       zMinistralModel.forward)NNNNNNN)rErFrGrr@rrr rr LongTensorTensorr FloatTensorrMr r r rr^rCrCr\rDrcs>  rcc@rO)MinistralForCausalLMNrQrCrCrCrDrrRrc@rO)"MinistralForSequenceClassificationNrQrCrCrCrDrrRrc@rO)MinistralForTokenClassificationNrQrCrCrCrDrrRrc@rO)MinistralForQuestionAnsweringNrQrCrCrCrDrrRr)rrarcrrrr)2rrr cache_utilsrrconfiguration_utilsr masking_utilsrrmodeling_outputsr modeling_rope_utilsr processing_utilsr utilsr r utils.genericrutils.output_capturingrmistral.configuration_mistralrqwen2.modeling_qwen2rrrrrrrrrrrrrPrSr_r`rarbrcrrrr__all__rCrCrCrDs6        4  L