
    sg                     ,    d dl mZ  G d de      ZdgZy)   )PretrainedConfigc                   Z     e Zd ZdZdZdgZdddddd	d
ddddddddddg dddf fd	Z xZS )	GlmConfiga  
    This is the configuration class to store the configuration of a [`GlmModel`]. It is used to instantiate an Glm
    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 Glm-4-9b-chat.
    e.g. [THUDM/glm-4-9b-chat](https://huggingface.co/THUDM/glm-4-9b-chat)
    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 151552):
            Vocabulary size of the Glm model. Defines the number of different tokens that can be represented by the
            `inputs_ids` passed when calling [`GlmModel`]
        hidden_size (`int`, *optional*, defaults to 4096):
            Dimension of the hidden representations.
        intermediate_size (`int`, *optional*, defaults to 13696):
            Dimension of the MLP representations.
        num_hidden_layers (`int`, *optional*, defaults to 40):
            Number of hidden layers in the Transformer decoder.
        num_attention_heads (`int`, *optional*, defaults to 32):
            Number of attention heads for each attention layer in the Transformer decoder.
        num_key_value_heads (`int`, *optional*, defaults to 2):
            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 checkout [this
            paper](https://arxiv.org/pdf/2305.13245.pdf). If it is not specified, will default to
            `num_attention_heads`.
        partial_rotary_factor (`float`, *optional*, defaults to 0.5): The factor of the partial rotary position.
        head_dim (`int`, *optional*, defaults to 128):
            The attention head dimension.
        hidden_act (`str` or `function`, *optional*, defaults to `"silu"`):
            The legacy activation function. It is overwritten by the `hidden_activation`.
        attention_dropout (`float`, *optional*, defaults to 0.0):
            The dropout ratio for the attention probabilities.
        max_position_embeddings (`int`, *optional*, defaults to 131072):
            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.
        rms_norm_eps (`float`, *optional*, defaults to 1.5625e-07):
            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`.
        tie_word_embeddings (`bool`, *optional*, defaults to `False`):
            Whether to tie weight embeddings
        rope_theta (`float`, *optional*, defaults to 10000.0):
            The base period of the RoPE embeddings.
        pad_token_id (`int`, *optional*, defaults to 151329):
            Padding token id.
        eos_token_id (`int` | `list`, *optional*, defaults to `[151329, 151336, 151338]`):
            End of stream token id.
        bos_token_id (`int`, *optional*):
            Beginning of stream token id.
        attention_bias (`bool`, defaults to `False`, *optional*, defaults to `True`):
            Whether to use a bias in the query, key, value and output projection layers during self-attention.
    ```python
    >>> from transformers import GlmModel, GlmConfig
    >>> # Initializing a Glm glm-4-9b-chat style configuration
    >>> configuration = GlmConfig()
    >>> # Initializing a model from the glm-4-9b-chat style configuration
    >>> model = GlmModel(configuration)
    >>> # Accessing the model configuration
    >>> configuration = model.config
    ```glmpast_key_valuesi P i   i5  (          g      ?   silug        i   g{Gz?gh㈵>TFg     @!O )r   i(O i*O Nc                    || _         || _        || _        || _        || _        || _        || _        || _        || _        |	| _	        || _
        || _        || _        || _        || _        |
| _        t!        | D  d||||d| y )N)pad_token_idbos_token_ideos_token_idtie_word_embeddings )
vocab_sizemax_position_embeddingshidden_sizeintermediate_sizenum_hidden_layersnum_attention_headspartial_rotary_factorhead_dimnum_key_value_heads
hidden_actinitializer_rangerms_norm_eps	use_cache
rope_thetaattention_biasattention_dropoutsuper__init__)selfr   r   r   r   r   r   r   r   r   r#   r   r   r   r    r   r!   r   r   r   r"   kwargs	__class__s                         \/var/www/html/venv/lib/python3.12/site-packages/transformers/models/glm/configuration_glm.pyr%   zGlmConfig.__init__Y   s    0 %'>$&!2!2#6 %:" #6 $!2("$,!2 	
%%% 3		

 	
    )__name__
__module____qualname____doc__
model_typekeys_to_ignore_at_inferencer%   __classcell__)r(   s   @r)   r   r      sa    ?B J#4"5 ! &"!-+/
 /
r*   r   N)configuration_utilsr   r   __all__r   r*   r)   <module>r4      s#   " 4t
  t
n -r*   