gLdZddlmZddlmZddlmZddlmZddl m Z e j e Z Gdd eZGd d eZd S) zSqueezeBERT model configuration) OrderedDict)Mapping)PretrainedConfig) OnnxConfig)loggingcPeZdZdZdZ dfd ZxZS)SqueezeBertConfiga This is the configuration class to store the configuration of a [`SqueezeBertModel`]. It is used to instantiate a SqueezeBERT 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 SqueezeBERT [squeezebert/squeezebert-uncased](https://huggingface.co/squeezebert/squeezebert-uncased) 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 30522): Vocabulary size of the SqueezeBERT model. Defines the number of different tokens that can be represented by the `inputs_ids` passed when calling [`SqueezeBertModel`]. hidden_size (`int`, *optional*, defaults to 768): Dimensionality of the encoder layers and the pooler layer. num_hidden_layers (`int`, *optional*, defaults to 12): Number of hidden layers in the Transformer encoder. num_attention_heads (`int`, *optional*, defaults to 12): Number of attention heads for each attention layer in the Transformer encoder. intermediate_size (`int`, *optional*, defaults to 3072): Dimensionality of the "intermediate" (often named feed-forward) layer in the Transformer encoder. hidden_act (`str` or `Callable`, *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. hidden_dropout_prob (`float`, *optional*, defaults to 0.1): The dropout probability for all fully connected layers in the embeddings, encoder, and pooler. attention_probs_dropout_prob (`float`, *optional*, defaults to 0.1): The dropout ratio for the attention probabilities. max_position_embeddings (`int`, *optional*, defaults to 512): 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). type_vocab_size (`int`, *optional*, defaults to 2): The vocabulary size of the `token_type_ids` passed when calling [`BertModel`] or [`TFBertModel`]. initializer_range (`float`, *optional*, defaults to 0.02): The standard deviation of the truncated_normal_initializer for initializing all weight matrices. layer_norm_eps (`float`, *optional*, defaults to 1e-12): pad_token_id (`int`, *optional*, defaults to 0): The ID of the token in the word embedding to use as padding. embedding_size (`int`, *optional*, defaults to 768): The dimension of the word embedding vectors. q_groups (`int`, *optional*, defaults to 4): The number of groups in Q layer. k_groups (`int`, *optional*, defaults to 4): The number of groups in K layer. v_groups (`int`, *optional*, defaults to 4): The number of groups in V layer. post_attention_groups (`int`, *optional*, defaults to 1): The number of groups in the first feed forward network layer. intermediate_groups (`int`, *optional*, defaults to 4): The number of groups in the second feed forward network layer. output_groups (`int`, *optional*, defaults to 4): The number of groups in the third feed forward network layer. Examples: ```python >>> from transformers import SqueezeBertConfig, SqueezeBertModel >>> # Initializing a SqueezeBERT configuration >>> configuration = SqueezeBertConfig() >>> # Initializing a model (with random weights) from the configuration above >>> model = SqueezeBertModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ``` squeezebert:w gelu皙?{Gz?-q=rc Htjdd| i|||_||_||_||_||_||_||_||_ | |_ | |_ | |_ | |_ ||_||_||_||_||_||_||_dS)N pad_token_id)super__init__ vocab_size hidden_sizenum_hidden_layersnum_attention_heads hidden_actintermediate_sizehidden_dropout_probattention_probs_dropout_probmax_position_embeddingstype_vocab_sizeinitializer_rangelayer_norm_epsembedding_sizeq_groupsk_groupsv_groupspost_attention_groupsintermediate_groups output_groups)selfrrrr r"r!r#r$r%r&r'r(rr)r*r+r,r-r.r/kwargs __class__s u/var/www/html/ai-engine/env/lib/python3.11/site-packages/transformers/models/squeezebert/configuration_squeezebert.pyrzSqueezeBertConfig.__init__gs0 ==l=f===$&!2#6 $!2#6 ,H)'>$.!2,,      %:"#6 *)r r rrrrrrrrrrrr rrrrrr)__name__ __module__ __qualname____doc__ model_typer __classcell__)r2s@r3r r sFFPJ%( #+,+,+,+,+,+,+,+,+,+,+r4r cJeZdZedeeeeefffdZdS)SqueezeBertOnnxConfigreturnc`|jdkrdddd}nddd}td|fd|fd |fgS) Nzmultiple-choicebatchchoicesequence)rrr)rr input_idsattention_masktoken_type_ids)taskr)r0 dynamic_axiss r3inputszSqueezeBertOnnxConfig.inputss] 9) ) )&8 CCLL&:66Ll+!<0!<0    r4N)r5r6r7propertyrstrintrGrr4r3r<r<sL   WS#X%6 67    X    r4r<N)r8 collectionsrtypingrconfiguration_utilsronnxrutilsr get_loggerr5loggerr r<rr4r3rRs&%######333333  H % %w+w+w+w+w+(w+w+w+v      J      r4