
    fThQ%                         S r SSKJr  SSKJrJrJrJr  SSKJ	r	J
r
Jr  SSKJr  SSKJrJr  SSKJr  \R&                  " \5      r " S	 S
\5      r " S S\5      rS
S/rg)zCodeGen model configuration    )OrderedDict)AnyListMappingOptional   )PreTrainedTokenizer
TensorTypeis_torch_available)PretrainedConfig)OnnxConfigWithPastPatchingSpec)loggingc                   f   ^  \ rS rSrSrSrSSSSS.r                  SU 4S	 jjrS
rU =r	$ )CodeGenConfig   a  
This is the configuration class to store the configuration of a [`CodeGenModel`]. It is used to instantiate a
CodeGen 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 CodeGen
[Salesforce/codegen-2B-mono](https://huggingface.co/Salesforce/codegen-2B-mono) 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 50400):
        Vocabulary size of the CodeGen model. Defines the number of different tokens that can be represented by the
        `inputs_ids` passed when calling [`CodeGenModel`].
    n_positions (`int`, *optional*, defaults to 2048):
        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).
    n_ctx (`int`, *optional*, defaults to 2048):
        This attribute is used in `CodeGenModel.__init__` without any real effect.
    n_embd (`int`, *optional*, defaults to 4096):
        Dimensionality of the embeddings and hidden states.
    n_layer (`int`, *optional*, defaults to 28):
        Number of hidden layers in the Transformer encoder.
    n_head (`int`, *optional*, defaults to 16):
        Number of attention heads for each attention layer in the Transformer encoder.
    rotary_dim (`int`, *optional*, defaults to 64):
        Number of dimensions in the embedding that Rotary Position Embedding is applied to.
    n_inner (`int`, *optional*):
        Dimensionality of the inner feed-forward layers. `None` will set it to 4 times n_embd
    activation_function (`str`, *optional*, defaults to `"gelu_new"`):
        Activation function, to be selected in the list `["relu", "silu", "gelu", "tanh", "gelu_new"]`.
    resid_pdrop (`float`, *optional*, defaults to 0.0):
        The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
    embd_pdrop (`int`, *optional*, defaults to 0.0):
        The dropout ratio for the embeddings.
    attn_pdrop (`float`, *optional*, defaults to 0.0):
        The dropout ratio for the attention.
    layer_norm_epsilon (`float`, *optional*, defaults to 1e-05):
        The epsilon to use in the layer normalization layers.
    initializer_range (`float`, *optional*, defaults to 0.02):
        The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
    use_cache (`bool`, *optional*, defaults to `True`):
        Whether or not the model should return the last key/values attentions (not used by all models).
    bos_token_id (`int`, *optional*, defaults to 50256):
        Beginning of stream token id.
    eos_token_id (`int`, *optional*, defaults to 50256):
        End of stream token id.
    tie_word_embeddings (`bool`, *optional*, defaults to `False`):
        Whether the model's input and output word embeddings should be tied. Note that this is only relevant if the
        model has a output word embedding layer.

Example:

```python
>>> from transformers import CodeGenConfig, CodeGenModel

>>> # Initializing a CodeGen 6B configuration
>>> configuration = CodeGenConfig()

>>> # Initializing a model (with random weights) from the configuration
>>> model = CodeGenModel(configuration)

>>> # Accessing the model configuration
>>> configuration = model.config
```codegenn_positionsn_embdn_headn_layer)max_position_embeddingshidden_sizenum_attention_headsnum_hidden_layersc                    > Xl         X0l        X l        X@l        XPl        X`l        Xl        Xpl        Xl        Xl	        Xl
        Xl        Xl        Xl        Xl        UU l        UU l        ["        TU ]H  " SUUUS.UD6  g )N)bos_token_ideos_token_idtie_word_embeddings )
vocab_sizen_ctxr   r   r   r   n_inner
rotary_dimactivation_functionresid_pdrop
embd_pdrop
attn_pdroplayer_norm_epsiloninitializer_range	use_cacher   r   super__init__)selfr!   r   r"   r   r   r   r$   r#   r%   r&   r'   r(   r)   r*   r+   r   r   r   kwargs	__class__s                       i/var/www/auris/envauris/lib/python3.13/site-packages/transformers/models/codegen/configuration_codegen.pyr-   CodeGenConfig.__init__f   s    , %
&$#6 &$$"4!2"(( 	
%LVi	
ms	
    )r%   r(   r   r'   r   r*   r)   r"   r   r   r#   r   r   r&   r$   r+   r!   )i     r4   i         @   Ngelu_new        r9   r9   gh㈵>g{Gz?TP  r:   F)
__name__
__module____qualname____firstlineno____doc__
model_typeattribute_mapr-   __static_attributes____classcell__r0   s   @r1   r   r      se    >@ J#0'&	M &!'+
 +
r3   r   c                   
  ^  \ rS rSr   SS\S\S\\\      S\	4U 4S jjjr
\S\\\\\4   4   4S j5       r\S\4S	 j5       r\S\4S
 j5       r    SS\S\S\S\	S\\   S\\\4   4U 4S jjjr\S\4S j5       rSrU =r$ )CodeGenOnnxConfig   configtaskpatching_specsuse_pastc                 ~   > [         TU ]  XX4S9  [        U R                  SS 5      (       d  SU R                  l        g g )N)rI   rJ   rK   pad_token_idr   )r,   r-   getattr_configrM   )r.   rH   rI   rJ   rK   r0   s        r1   r-   CodeGenOnnxConfig.__init__   s;     	>]t||^T::()DLL% ;r3   returnc                     [        SSSS.05      nU R                  (       a  U R                  USS9  SSS.US'   U$ SSS.US'   U$ )	N	input_idsbatchsequence)r      inputs)	directionzpast_sequence + sequenceattention_mask)r   rK   fill_with_past_key_values_)r.   common_inputss     r1   rW   CodeGenOnnxConfig.inputs   sa    #[g*2M$NO==++MX+N29>X.YM*+  3:j.IM*+r3   c                 .    U R                   R                  $ N)rO   r   r.   s    r1   
num_layersCodeGenOnnxConfig.num_layers   s    ||###r3   c                 .    U R                   R                  $ r^   )rO   r   r_   s    r1   r   %CodeGenOnnxConfig.num_attention_heads   s    ||"""r3   	tokenizer
batch_size
seq_lengthis_pair	frameworkc           	      r  > [         [        U ]  XX4US9n[        SUS   05      nU R                  (       a  [        5       (       d  [        S5      eSS KnUS   R                  u  pU
S-   nU	U R                  UU R                  R                  U R                  -  4n[        U R                  5       Vs/ s H$  oR                  U5      UR                  U5      4PM&     snUS'   US   US'   U R                  (       a6  US   R                  nWR!                  US   UR#                  W	WUS9/S	S
9US'   U$ s  snf )N)re   rf   rg   rh   rS   zACannot generate dummy past_keys inputs without PyTorch installed.r      past_key_valuesrY   )dtyperV   )dim)r,   r   generate_dummy_inputsr   rK   r   
ValueErrortorchshaper   rO   r   ranger`   zerosrl   catones)r.   rd   re   rf   rg   rh   r[   ordered_inputsrp   rT   seqlenpast_key_values_length
past_shape_
mask_dtyper0   s                  r1   rn   'CodeGenOnnxConfig.generate_dummy_inputs   s^    0$M`i N 

 %k=3M%NO ==%'' !dee -k : @ @)/!&,,*LL,,0H0HH	
 QVVZVeVePf5Pf1[[,ekk*.EFPf501 ,99I+J'(=='(89??J/4yy 015::eE[cm:3nouv 09 0N+, 5s   2+D4c                     g)N   r    r_   s    r1   default_onnx_opset$CodeGenOnnxConfig.default_onnx_opset   s    r3   r    )defaultNF)r   FN)r;   r<   r=   r>   r   strr   r   r   boolr-   propertyr   intrW   r`   r   r	   r
   r   rn   r   rB   rC   rD   s   @r1   rF   rF      s(    7;
* 
* 
* !l!34	
*
 
* 
* WS#X%6 67   $C $ $ #S # # *.*&* * 	*
 * J'* 
c	* *X C  r3   rF   N)r?   collectionsr   typingr   r   r   r    r	   r
   r   configuration_utilsr   onnxr   r   utilsr   
get_loggerr;   loggerr   rF   __all__r    r3   r1   <module>r      sa    " # / / C C 3 4  
		H	%t
$ t
pN* Nb /
0r3   