qiy`dZddlmZddlmZej eZGddeZdgZ y)zXGLM model configuration)PreTrainedConfig)loggingcbeZdZdZdZdgZddddZ d fd ZxZS) XGLMConfiga# This is the configuration class to store the configuration of a [`XGLMModel`]. It is used to instantiate an XGLM 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 XGLM [facebook/xglm-564M](https://huggingface.co/facebook/xglm-564M) 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 256008): Vocabulary size of the XGLM model. Defines the number of different tokens that can be represented by the `inputs_ids` passed when calling [`XGLMModel`]. max_position_embeddings (`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). d_model (`int`, *optional*, defaults to 1024): Dimension of the layers and the pooler layer. ffn_dim (`int`, *optional*, defaults to 4096): Dimension of the "intermediate" (often named feed-forward) layer in decoder. num_layers (`int`, *optional*, defaults to 24): Number of hidden layers Transformer decoder. attention_heads (`int`, *optional*, defaults to 16): Number of attention heads for each attention layer in the Transformer 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, dencoder, and pooler. attention_dropout (`float`, *optional*, defaults to 0.1): 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. layerdrop (`float`, *optional*, defaults to 0.0): The LayerDrop probability for the encoder. See the [LayerDrop paper](see https://huggingface.co/papers/1909.11556) for more details. init_std (`float`, *optional*, defaults to 0.02): The standard deviation of the truncated_normal_initializer for initializing all weight matrices. scale_embedding (`bool`, *optional*, defaults to `True`): 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). Example: ```python >>> from transformers import XGLMModel, XGLMConfig >>> # Initializing a XGLM facebook/xglm-564M style configuration >>> configuration = XGLMConfig() >>> # Initializing a model from the facebook/xglm-564M style configuration >>> model = XGLMModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ```xglmpast_key_valuesattention_headsd_model num_layers)num_attention_heads hidden_sizenum_hidden_layersc <||_||_||_||_||_||_||_||_||_| |_ | |_ | |_ | |_ | |_ ||_||_||_||_||_||_t)|Tdi|y)N)add_cross_attention vocab_sizemax_position_embeddingsr ffn_dimr r activation_functiondropoutattention_dropoutactivation_dropout layerdropinit_stdscale_embedding use_cache pad_token_id bos_token_id eos_token_idtie_word_embeddingsdecoder_start_token_idsuper__init__)selfrrr rr r rrrrrrrrr!rrrrr kwargs __class__s ]/opt/pipecat/venv/lib/python3.12/site-packages/transformers/models/xglm/configuration_xglm.pyr#zXGLMConfig.__init__\s0$7 $'>$  $.#6  !2"4"  ."(((#6 &<# "6")iiiigelu皙?r,r-g{Gz?TTr.FT) __name__ __module__ __qualname____doc__ model_typekeys_to_ignore_at_inference attribute_mapr# __classcell__)r&s@r'rrss9 vJ#4"5 1 )M $" ! +-#-#r(rN) r4configuration_utilsrutilsr get_loggerr1loggerr__all__rr(r'r>s=3   H %r#!r#j .r(