o
    h2J                     @   s   d Z ddlZddlmZ ddlmZmZmZ ddlm	Z	 ddl
mZ ddlmZmZmZ dd	lmZ dd
lmZmZmZ eeZddiZG dd deZG dd deZdS )z BART model configuration    N)OrderedDict)AnyMappingOptional   )PreTrainedTokenizer)PretrainedConfig)
OnnxConfigOnnxConfigWithPastOnnxSeq2SeqConfigWithPast) compute_effective_axis_dimension)
TensorTypeis_torch_availableloggingzfacebook/bart-largezChttps://huggingface.co/facebook/bart-large/resolve/main/config.jsonc                       sj   e Zd ZdZdZdgZdddZ					
				
																			d fdd	Z  ZS )
BartConfiga  
    This is the configuration class to store the configuration of a [`BartModel`]. It is used to instantiate a BART
    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 BART
    [facebook/bart-large](https://huggingface.co/facebook/bart-large) architecture.

    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 50265):
            Vocabulary size of the BART model. Defines the number of different tokens that can be represented by the
            `inputs_ids` passed when calling [`BartModel`] or [`TFBartModel`].
        d_model (`int`, *optional*, defaults to 1024):
            Dimensionality of the layers and the pooler layer.
        encoder_layers (`int`, *optional*, defaults to 12):
            Number of encoder layers.
        decoder_layers (`int`, *optional*, defaults to 12):
            Number of decoder layers.
        encoder_attention_heads (`int`, *optional*, defaults to 16):
            Number of attention heads for each attention layer in the Transformer encoder.
        decoder_attention_heads (`int`, *optional*, defaults to 16):
            Number of attention heads for each attention layer in the Transformer decoder.
        decoder_ffn_dim (`int`, *optional*, defaults to 4096):
            Dimensionality of the "intermediate" (often named feed-forward) layer in decoder.
        encoder_ffn_dim (`int`, *optional*, defaults to 4096):
            Dimensionality of the "intermediate" (often named feed-forward) layer in decoder.
        activation_function (`str` or `function`, *optional*, defaults to `"gelu"`):
            The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`,
            `"relu"`, `"silu"` and `"gelu_new"` are supported.
        dropout (`float`, *optional*, defaults to 0.1):
            The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
        attention_dropout (`float`, *optional*, defaults to 0.0):
            The dropout ratio for the attention probabilities.
        activation_dropout (`float`, *optional*, defaults to 0.0):
            The dropout ratio for activations inside the fully connected layer.
        classifier_dropout (`float`, *optional*, defaults to 0.0):
            The dropout ratio for classifier.
        max_position_embeddings (`int`, *optional*, defaults to 1024):
            The maximum sequence length that this model might ever be used with. Typically set this to something large
            just in case (e.g., 512 or 1024 or 2048).
        init_std (`float`, *optional*, defaults to 0.02):
            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
        encoder_layerdrop (`float`, *optional*, defaults to 0.0):
            The LayerDrop probability for the encoder. See the [LayerDrop paper](see https://arxiv.org/abs/1909.11556)
            for more details.
        decoder_layerdrop (`float`, *optional*, defaults to 0.0):
            The LayerDrop probability for the decoder. See the [LayerDrop paper](see https://arxiv.org/abs/1909.11556)
            for more details.
        scale_embedding (`bool`, *optional*, defaults to `False`):
            Scale embeddings by diving by sqrt(d_model).
        use_cache (`bool`, *optional*, defaults to `True`):
            Whether or not the model should return the last key/values attentions (not used by all models).
        num_labels (`int`, *optional*, defaults to 3):
            The number of labels to use in [`BartForSequenceClassification`].
        forced_eos_token_id (`int`, *optional*, defaults to 2):
            The id of the token to force as the last generated token when `max_length` is reached. Usually set to
            `eos_token_id`.

    Example:

    ```python
    >>> from transformers import BartConfig, BartModel

    >>> # Initializing a BART facebook/bart-large style configuration
    >>> configuration = BartConfig()

    >>> # Initializing a model (with random weights) from the facebook/bart-large style configuration
    >>> model = BartModel(configuration)

    >>> # Accessing the model configuration
    >>> configuration = model.config
    ```bartpast_key_valuesencoder_attention_headsd_model)num_attention_headshidden_sizeY                      gelu皙?{Gz?FTr      r      c              
      s   || _ || _|| _|| _|| _|| _|| _|| _|| _|| _	|| _
|| _|| _|| _|	| _|
| _|| _|| _|| _|| _t jd|||||||d| | jd u rh|ddrj| j| _td| j d d S d S d S )N)
num_labelspad_token_idbos_token_ideos_token_idis_encoder_decoderdecoder_start_token_idforced_eos_token_idforce_bos_token_to_be_generatedFz:Please make sure the config includes `forced_bos_token_id=zT` in future versions. The config can simply be saved and uploaded again to be fixed. )
vocab_sizemax_position_embeddingsr   encoder_ffn_dimencoder_layersr   decoder_ffn_dimdecoder_layersdecoder_attention_headsdropoutattention_dropoutactivation_dropoutactivation_functioninit_stdencoder_layerdropdecoder_layerdropclassifier_dropout	use_cachenum_hidden_layersscale_embeddingsuper__init__forced_bos_token_idgetr$   warningswarn)selfr+   r,   r.   r-   r   r0   r/   r1   r7   r8   r5   r   r2   r3   r4   r6   r9   r<   r:   r"   r#   r$   r%   r&   r'   r(   kwargs	__class__r*   a/var/www/html/ai/venv/lib/python3.10/site-packages/transformers/models/bart/configuration_bart.pyr>   s   sJ   zBartConfig.__init__)r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   FTr   r    r   r!   Tr!   r!   )	__name__
__module____qualname____doc__
model_typekeys_to_ignore_at_inferenceattribute_mapr>   __classcell__r*   r*   rE   rG   r   #   s@    K
r   c                       sD  e Zd Zedeeeeef f fddZedeeeeef f f fddZ				dd	e	d
edede
dee deeef fddZ				dd	e	d
edede
dee deeef fddZ				dd	e	d
edede
dee deeef fddZ				dd	e	d
edede
dee deeef fddZ fddZ  ZS )BartOnnxConfigreturnc                 C   s4  | j dv r@tddddfddddfg}| jr&ddi|d< dd	d|d
< nddd|d< ddd|d
< | jr>| j|dd |S | j dkr|tddddfddddfg}| jrz| j\}}t|D ]}ddd|d| d< ddd|d| d< qa|S tddddfddddfddddfd
dddfg}|S )Ndefaultz
seq2seq-lm	input_idsbatchencoder_sequence)r   r    attention_maskr   decoder_input_idsz past_decoder_sequence + sequencedecoder_attention_maskdecoder_sequenceinputs)	direction	causal-lmpast_sequence + sequencer   r!   zpast_key_values..key.value)taskr   use_pastfill_with_past_key_values_
num_layersrange)rC   common_inputsnum_encoder_layers_ir*   r*   rG   r[      sD   


	zBartOnnxConfig.inputsc                    sp   | j dv rt j}|S tt| j}| jr6| j\}}t|D ]}ddd|d| d< ddd|d| d< q|S )NrR   rU   r^   r_   zpresent.r`   ra   )rb   r=   outputsr
   rc   re   rf   )rC   common_outputsrh   ri   rj   rE   r*   rG   rk      s   

zBartOnnxConfig.outputsFN	tokenizer
batch_size
seq_lengthis_pair	frameworkc              	   C   s  |  |||||}| js|nd}|  |||||}dd | D }tdi ||}	| jrt s5tddd l}
|	d j\}}|	d jd }| j\}}|||| j	j
| f}|d }|||| j	j
| f}|
j|	d	 |
||gdd
|	d	< g |	d< | j\}}t||}t||| }||krdnd}t|D ]}|	d |
||
||
||
|f q|dkr|n|}t||D ]}|	d |
||
|f q|	S )Nr    c                 S   s   i | ]
\}}d | |qS )decoder_r*   ).0nametensorr*   r*   rG   
<dictcomp>  s    zTBartOnnxConfig._generate_dummy_inputs_for_default_and_seq2seq_lm.<locals>.<dictcomp>ACannot generate dummy past_keys inputs without PyTorch installed.r   rT   rX   r   rY   dimr   encoderdecoderr*   )I_generate_dummy_inputs_for_sequence_classification_and_question_answeringrc   itemsdictr   
ValueErrortorchshaper   _configr   catonesre   minmaxrf   appendzeros)rC   rn   ro   rp   rq   rr   encoder_inputsdecoder_seq_lengthdecoder_inputsrg   r   rU   encoder_seq_lengthnum_encoder_attention_headsnum_decoder_attention_headsencoder_shapedecoder_past_lengthdecoder_shaperh   num_decoder_layersmin_num_layersmax_num_layersremaining_side_nameri   r   r*   r*   rG   1_generate_dummy_inputs_for_default_and_seq2seq_lm   s^   







	 z@BartOnnxConfig._generate_dummy_inputs_for_default_and_seq2seq_lmc                    s   |  |||||}| jr\t stddd l|d j\}}|d }	| j\}
}| j\}}|||	| jj	| f |d j
}j|d j||	|dgdd|d<  fd	d
t|
D |d< |S )Nrx   r   rT   r!   rW   )dtyper    ry   c                    s    g | ]}    fqS r*   )r   )rt   ri   
past_shaper   r*   rG   
<listcomp>Y  s    zGBartOnnxConfig._generate_dummy_inputs_for_causal_lm.<locals>.<listcomp>r   )r}   rc   r   r   r   r   re   r   r   r   r   r   r   rf   )rC   rn   ro   rp   rq   rr   rg   rU   seqlenpast_key_values_lengthrh   ri   r   
mask_dtyper*   r   rG   $_generate_dummy_inputs_for_causal_lm8  s0   






z3BartOnnxConfig._generate_dummy_inputs_for_causal_lmc           	      C   sV   t |tjdd}||}t |tj|d}d|jg| g| }t|||d}|S )Nr   )fixed_dimensionnum_token_to_add )return_tensors)r   r	   default_fixed_batchnum_special_tokens_to_adddefault_fixed_sequencejoin	unk_tokenr   )	rC   rn   ro   rp   rq   rr   token_to_adddummy_inputrg   r*   r*   rG   r}   ^  s   
zXBartOnnxConfig._generate_dummy_inputs_for_sequence_classification_and_question_answeringc                 C   s\   | j dv r| j|||||d}|S | j dkr"| j|||||d}|S | j|||||d}|S )NrR   )ro   rp   rq   rr   r]   )rb   r   r   r}   )rC   rn   ro   rp   rq   rr   rg   r*   r*   rG   generate_dummy_inputsx  s   




z$BartOnnxConfig.generate_dummy_inputsc                    s:   | j dv rt ||||}d S tt| ||||}d S )NrR   )rb   r=   _flatten_past_key_values_r   )rC   flattened_outputru   idxtrE   r*   rG   r     s
   

z(BartOnnxConfig._flatten_past_key_values_)rm   rm   FN)rH   rI   rJ   propertyr   strintr[   rk   r   boolr   r   r   r   r   r}   r   r   rO   r*   r*   rE   rG   rP      s     +$

G

)



rP   )rK   rA   collectionsr   typingr   r   r    r   configuration_utilsr   onnxr	   r
   r   
onnx.utilsr   utilsr   r   r   
get_loggerrH   logger"BART_PRETRAINED_CONFIG_ARCHIVE_MAPr   rP   r*   r*   r*   rG   <module>   s   
 