o
    iR                     @   sl  d dl mZmZmZ d dlZd dlmZ d dlmZ ddlm	Z	 ddl
mZmZ ddlmZ dd	lmZ dd
lmZ ddlmZmZmZmZ ddlmZmZ ddlmZmZ ddlmZm Z  ddl!m"Z" ddl#m$Z$m%Z% ddl&m'Z' ddl(m)Z) ddl*m+Z+ G dd dej,Z-edG dd dej,Z.G dd dej,Z/dd Z0dCddZ1d ej2d!e3d"ej2fd#d$Z4	%dDd&ej,d'ej2d(ej2d)ej2d*eej2 d+e5d,e5d-e"e$ fd.d/Z6G d0d1 d1ej,Z7G d2d3 d3eZ8eG d4d5 d5e Z9eG d6d7 d7e9Z:ed8d9G d:d; d;e9eZ;ed8d9G d<d= d=ee9Z<ed8d9G d>d? d?ee9Z=ed8d9G d@dA dAee9Z>g dBZ?dS )E    )CallableOptionalUnionN)nn)auto_docstring   )ACT2FN)CacheDynamicCache)GenerationMixin)use_kernel_forward_from_hub)create_causal_mask)GenericForQuestionAnswering GenericForSequenceClassificationGenericForTokenClassificationGradientCheckpointingLayer)BaseModelOutputWithPastCausalLMOutputWithPast)ROPE_INIT_FUNCTIONSdynamic_rope_update)ALL_ATTENTION_FUNCTIONSPreTrainedModel)Unpack)TransformersKwargscan_return_tuple)deprecate_kwarg)check_model_inputs   )ArceeConfigc                       s$   e Zd Z fddZdd Z  ZS )ArceeMLPc                    s`   t    || _|j| _|j| _tj| j| j|jd| _tj| j| j|jd| _	t
|j | _d S )Nbias)super__init__confighidden_sizeintermediate_sizer   Linearmlp_biasup_proj	down_projr   
hidden_actact_fnselfr$   	__class__ \/home/ubuntu/.local/lib/python3.10/site-packages/transformers/models/arcee/modeling_arcee.pyr#   3   s   
zArceeMLP.__init__c                 C   s   |  | | |S N)r*   r,   r)   )r.   xr1   r1   r2   forward<   s   zArceeMLP.forward)__name__
__module____qualname__r#   r5   __classcell__r1   r1   r/   r2   r   2   s    	r   RMSNormc                       s.   e Zd Zd fdd	Zdd Zdd Z  ZS )	ArceeRMSNormư>c                    s&   t    tt|| _|| _dS )z;
        ArceeRMSNorm is equivalent to T5LayerNorm
        N)r"   r#   r   	Parametertorchonesweightvariance_epsilon)r.   r%   epsr/   r1   r2   r#   B   s   

zArceeRMSNorm.__init__c                 C   sJ   |j }|tj}|djddd}|t|| j  }| j|| S )N   T)keepdim)	dtypetor>   float32powmeanrsqrtrA   r@   )r.   hidden_statesinput_dtypevariancer1   r1   r2   r5   J   s
   zArceeRMSNorm.forwardc                 C   s   t | jj d| j S )Nz, eps=)tupler@   shaperA   )r.   r1   r1   r2   
extra_reprQ   s   zArceeRMSNorm.extra_repr)r<   )r6   r7   r8   r#   r5   rQ   r9   r1   r1   r/   r2   r;   @   s    r;   c                       sD   e Zd ZU ejed< ddef fddZe e	dd Z
  ZS )	ArceeRotaryEmbeddinginv_freqNr$   c                    s   t    t|drt|jtr|jd|jd| _nd| _|j| _	|j| _
|| _t| j | _| | j|\}| _| jd|dd | j| _d S )Nrope_scaling	rope_typetypedefaultrS   F)
persistent)r"   r#   hasattr
isinstancerT   dictgetrU   max_position_embeddingsmax_seq_len_cachedoriginal_max_seq_lenr$   r   rope_init_fnattention_scalingregister_bufferrS   original_inv_freq)r.   r$   devicerS   r/   r1   r2   r#   X   s   
zArceeRotaryEmbedding.__init__c           
      C   s   | j d d d d f  |jd dd|j}|d d d d d f  }t|jjtr6|jjdkr6|jjnd}t	j
|dd+ | |  dd}t	j||fdd	}| | j }| | j }	W d    n1 smw   Y  |j|jd
|	j|jd
fS )Nr   rD   r   mpscpuF)device_typeenabledrC   dim)rF   )rS   floatexpandrP   rG   rd   rZ   rV   strr>   autocast	transposecatcosra   sinrF   )
r.   r4   position_idsinv_freq_expandedposition_ids_expandedrg   freqsembrq   rr   r1   r1   r2   r5   i   s   0&zArceeRotaryEmbedding.forwardr3   )r6   r7   r8   r>   Tensor__annotations__r   r#   no_gradr   r5   r9   r1   r1   r/   r2   rR   U   s   
 
rR   c                 C   sH   | dd| j d d f }| d| j d d df }tj| |fddS )z*Rotates half the hidden dims of the input..NrD   rC   ri   )rP   r>   rp   )r4   x1x2r1   r1   r2   rotate_halfy   s   r}   c                 C   sD   | |}| |}| | t| |  }|| t||  }||fS )a  Applies Rotary Position Embedding to the query and key tensors.

    Args:
        q (`torch.Tensor`): The query tensor.
        k (`torch.Tensor`): The key tensor.
        cos (`torch.Tensor`): The cosine part of the rotary embedding.
        sin (`torch.Tensor`): The sine part of the rotary embedding.
        position_ids (`torch.Tensor`, *optional*):
            Deprecated and unused.
        unsqueeze_dim (`int`, *optional*, defaults to 1):
            The 'unsqueeze_dim' argument specifies the dimension along which to unsqueeze cos[position_ids] and
            sin[position_ids] so that they can be properly broadcasted to the dimensions of q and k. For example, note
            that cos[position_ids] and sin[position_ids] have the shape [batch_size, seq_len, head_dim]. Then, if q and
            k have the shape [batch_size, heads, seq_len, head_dim], then setting unsqueeze_dim=1 makes
            cos[position_ids] and sin[position_ids] broadcastable to the shapes of q and k. Similarly, if q and k have
            the shape [batch_size, seq_len, heads, head_dim], then set unsqueeze_dim=2.
    Returns:
        `tuple(torch.Tensor)` comprising of the query and key tensors rotated using the Rotary Position Embedding.
    )	unsqueezer}   )qkrq   rr   rs   unsqueeze_dimq_embedk_embedr1   r1   r2   apply_rotary_pos_emb   s
   

r   rL   n_repreturnc                 C   s^   | j \}}}}|dkr| S | dddddddddf |||||} | ||| ||S )z
    This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch,
    num_key_value_heads, seqlen, head_dim) to (batch, num_attention_heads, seqlen, head_dim)
    r   N)rP   rl   reshape)rL   r   batchnum_key_value_headsslenhead_dimr1   r1   r2   	repeat_kv   s
   0r           modulequerykeyvalueattention_maskscalingdropoutkwargsc                 K   s   t || j}t || j}	t||dd| }
|d ur3|d d d d d d d |jd f }|
| }
tjj|
dtj	d
|j}
tjj|
|| jd}
t|
|	}|dd }||
fS )NrC   r   rD   )rj   rF   )ptrainingr   )r   num_key_value_groupsr>   matmulro   rP   r   
functionalsoftmaxrH   rG   rF   r   r   
contiguous)r   r   r   r   r   r   r   r   
key_statesvalue_statesattn_weightscausal_maskattn_outputr1   r1   r2   eager_attention_forward   s   
&r   c                       s   e Zd ZdZdedef fddZedddd		
	
ddej	de
ej	ej	f deej	 dee deej dee de
ej	ej	f fddZ  ZS )ArceeAttentionz=Multi-headed attention from 'Attention Is All You Need' paperr$   	layer_idxc                    s   t    || _|| _t|d|j|j | _|j|j | _	| jd | _
|j| _d| _tj|j|j| j |jd| _tj|j|j| j |jd| _tj|j|j| j |jd| _tj|j| j |j|jd| _d S )Nr   g      Tr    )r"   r#   r$   r   getattrr%   num_attention_headsr   r   r   r   attention_dropout	is_causalr   r'   attention_biasq_projk_projv_projo_projr.   r$   r   r/   r1   r2   r#      s(   
zArceeAttention.__init__past_key_valuepast_key_values4.58new_nameversionNrL   position_embeddingsr   cache_positionr   r   c                 K   s$  |j d d }g |d| jR }| ||dd}	| ||dd}
| ||dd}|\}}t|	|
||\}	}
|d urW|||d}||
|| j	|\}
}t
}| jjdkret| jj }|| |	|
||f| jsqdn| j| jd|\}}|jg |dR   }| |}||fS )NrD   r   rC   )rr   rq   r   eagerr   )r   r   )rP   r   r   viewro   r   r   r   updater   r   r$   _attn_implementationr   r   r   r   r   r   r   )r.   rL   r   r   r   r   r   input_shapehidden_shapequery_statesr   r   rq   rr   cache_kwargsattention_interfacer   r   r1   r1   r2   r5      s8   


zArceeAttention.forward)NN)r6   r7   r8   __doc__r   intr#   r   r>   rx   rO   r   r	   
LongTensorr   r   r5   r9   r1   r1   r/   r2   r      s*    r   c                       s   e Zd Zdedef fddZedddd							
				ddejde	ej de	ej
 de	e de	e de	ej
 de	eejejf  dee dejfddZ  ZS )ArceeDecoderLayerr$   r   c                    sR   t    |j| _t||d| _t|| _t|j|jd| _	t|j|jd| _
d S )N)r$   r   rB   )r"   r#   r%   r   	self_attnr   mlpr;   rms_norm_epsinput_layernormpost_attention_layernormr   r/   r1   r2   r#   	  s   

zArceeDecoderLayer.__init__r   r   r   r   NFrL   r   rs   	use_cacher   r   r   r   c              
   K   s^   |}	|  |}| jd|||||||d|\}}
|	| }|}	| |}| |}|	| }|S )N)rL   r   rs   r   r   r   r   r1   )r   r   r   r   )r.   rL   r   rs   r   r   r   r   r   residual_r1   r1   r2   r5     s&   




zArceeDecoderLayer.forward)NNNFNN)r6   r7   r8   r   r   r#   r   r>   rx   r   r   r	   boolrO   r   r   r5   r9   r1   r1   r/   r2   r     s8    
	
r   c                   @   sH   e Zd ZU eed< dZdZdgZdgZdZ	dZ
dZdZdZeedZdS )ArceePreTrainedModelr$   modelTr   r   )rL   
attentionsN)r6   r7   r8   r   ry   base_model_prefixsupports_gradient_checkpointing_no_split_modules_skip_keys_device_placement_supports_flash_attn_supports_sdpa_supports_flex_attn_can_compile_fullgraph_supports_attention_backendr   r   _can_record_outputsr1   r1   r1   r2   r   6  s   
 
r   c                       s   e Zd Zdef fddZee							ddeej	 deej
 deej	 dee d	eej d
eej	 dee dee defddZ  ZS )
ArceeModelr$   c                    s   t     j| _ j| _t j j| j| _t	 fddt
 jD | _t j jd| _t d| _d| _|   d S )Nc                    s   g | ]}t  |qS r1   )r   ).0r   r$   r1   r2   
<listcomp>R  s    z'ArceeModel.__init__.<locals>.<listcomp>r   r   F)r"   r#   pad_token_idpadding_idx
vocab_sizer   	Embeddingr%   embed_tokens
ModuleListrangenum_hidden_layerslayersr;   r   normrR   
rotary_embgradient_checkpointing	post_initr-   r/   r   r2   r#   K  s   zArceeModel.__init__N	input_idsr   rs   r   inputs_embedsr   r   r   r   c              	   K   s   |d u |d uA rt d|d u r| |}|r!|d u r!t| jd}|d u r=|d ur-| nd}	tj|	|	|jd  |jd}|d u rF|	d}t
| j|||||d}
|}| ||}| jd | jj D ]}||f|
||||d|}qb| |}t||dS )	Nz:You must specify exactly one of input_ids or inputs_embedsr   r   r   )rd   )r$   input_embedsr   r   r   rs   )r   rs   r   r   r   )last_hidden_stater   )
ValueErrorr   r
   r$   get_seq_lengthr>   arangerP   rd   r~   r   r   r   r   r   r   )r.   r   r   rs   r   r   r   r   r   past_seen_tokensr   rL   r   decoder_layerr1   r1   r2   r5   [  sP   

	

zArceeModel.forward)NNNNNNN)r6   r7   r8   r   r#   r   r   r   r>   r   rx   r	   FloatTensorr   r   r   r   r5   r9   r1   r1   r/   r2   r   I  s<    	
r   zarcee-ai/AFM-4.5B)
checkpointc                       s   e Zd ZdgZddiZddgdgfiZ fddZee										dd
e	e
j de	e
j de	e
j de	e de	e
j de	e
j de	e de	e
j deee
jf dee defddZ  ZS )ArceeForCausalLMzlm_head.weightlm_headcolwise_reprL   logitsc                    s@   t  | t|| _|j| _tj|j|jdd| _| 	  d S )NFr    )
r"   r#   r   r   r   r   r'   r%   r   r   r-   r/   r1   r2   r#     s
   
zArceeForCausalLM.__init__Nr   r   r   rs   r   r   labelsr   r   logits_to_keepr   r   c
              
   K   s   | j d|||||||d|
}|j}t|	trt|	 dn|	}| |dd|ddf }d}|durB| jd||| jjd|
}t	|||j
|j|jdS )a  
        Example:

        ```python
        >>> from transformers import AutoTokenizer, ArceeForCausalLM

        >>> model = ArceeForCausalLM.from_pretrained("meta-arcee/Arcee-2-7b-hf")
        >>> tokenizer = AutoTokenizer.from_pretrained("meta-arcee/Arcee-2-7b-hf")

        >>> prompt = "Hey, are you conscious? Can you talk to me?"
        >>> inputs = tokenizer(prompt, return_tensors="pt")

        >>> # Generate
        >>> generate_ids = model.generate(inputs.input_ids, max_length=30)
        >>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
        "Hey, are you conscious? Can you talk to me?\nI'm not conscious, but I can talk to you."
        ```)r   r   rs   r   r   r   r   N)r   r   r   )lossr   r   rL   r   r1   )r   r   rZ   r   slicer   loss_functionr$   r   r   r   rL   r   )r.   r   r   rs   r   r   r   r   r   r   r   outputsrL   slice_indicesr   r   r1   r1   r2   r5     s0    zArceeForCausalLM.forward)	NNNNNNNNr   )r6   r7   r8   _tied_weights_keys_tp_plan_pp_planr#   r   r   r   r>   r   rx   r	   r   r   r   r   r   r   r   r5   r9   r1   r1   r/   r2   r     sN    		
r   c                   @      e Zd ZdS )ArceeForSequenceClassificationNr6   r7   r8   r1   r1   r1   r2   r        r  c                   @   s   e Zd ZdZdS )ArceeForQuestionAnsweringtransformerN)r6   r7   r8   r   r1   r1   r1   r2   r
    s    r
  c                   @   r  )ArceeForTokenClassificationNr  r1   r1   r1   r2   r    r	  r  )r   r
  r  r  r   r   )Nr   )r   )@typingr   r   r   r>   r   transformers.utilsr   activationsr   cache_utilsr	   r
   
generationr   integrationsr   masking_utilsr   modeling_layersr   r   r   r   modeling_outputsr   r   modeling_rope_utilsr   r   modeling_utilsr   r   processing_utilsr   utilsr   r   utils.deprecationr   utils.genericr   configuration_arceer   Moduler   r;   rR   r}   r   rx   r   r   rk   r   r   r   r   r   r   r  r
  r  __all__r1   r1   r1   r2   <module>   st   $

G.NK