qiU*pdZddlmZmZddlmZddlmZeje Z GddeZ dgZ y)zAFMoE model configuration)PreTrainedConfiglayer_type_validation)RopeParameters)loggingc<eZdZdZdZdgZdgdgfddgdgfdgdgfdZ d(d ed zd ed zd ed zd ed zded zded zded zded zded zded zded zde d zde d zde d zde d zde d zde e ee fzd zded zded zded zde d zded zd ed zd!e d zd"e d zd#e d zd$e d zd%e d zd&e d zf:fd' ZxZS)) AfmoeConfiga This is the configuration class to store the configuration of a [`AfmoeModel`]. It is used to instantiate an AFMoE model according to the specified arguments, defining the model architecture. Instantiating a configuration with the defaults will yield a similar configuration to that of [arcee-ai/Trinity-Mini](https://huggingface.co/arcee-ai/Trinity-Mini). AFMoE is an Adaptive Feedforward MoE (Mixture of Experts) model with token-choice routing, shared experts, and a hybrid attention mechanism combining sliding window and full attention patterns. 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 200192): Vocabulary size of the AFMoE model. Defines the number of different tokens that can be represented by the `inputs_ids` passed when calling [`AfmoeModel`]. hidden_size (`int`, *optional*, defaults to 2048): Dimension of the hidden representations. intermediate_size (`int`, *optional*, defaults to 6144): Dimension of the dense MLP representations. moe_intermediate_size (`int`, *optional*, defaults to 1408): Intermediate size of the routed expert MLPs. num_hidden_layers (`int`, *optional*, defaults to 32): Number of hidden layers in the Transformer decoder. num_dense_layers (`int`, *optional*, defaults to 1): Number of initial dense layers before MoE layers begin. Layers with index < num_dense_layers will use standard dense MLPs instead of MoE. num_attention_heads (`int`, *optional*, defaults to 16): Number of attention heads for each attention layer in the Transformer decoder. num_key_value_heads (`int`, *optional*): 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, check out [this paper](https://huggingface.co/papers/2305.13245). If it is not specified, will default to `num_attention_heads`. head_dim (`int`, *optional*, defaults to 128): The dimension of each attention head. hidden_act (`str` or `function`, *optional*, defaults to `"silu"`): The non-linear activation function (function or string) in the MLP blocks. max_position_embeddings (`int`, *optional*, defaults to 16384): 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 1e-05): 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 the model's input and output word embeddings should be tied. rope_theta (`float`, *optional*, defaults to 10000.0): The base period of the RoPE embeddings. rope_parameters (`RopeParameters`, *optional*): Dictionary containing the configuration parameters for the RoPE embeddings. The dictionary should contain a value for `rope_theta` and optionally parameters used for scaling in case you want to use RoPE with longer `max_position_embeddings`. num_experts (`int`, *optional*, defaults to 64): Number of routed experts in MoE layers. num_experts_per_tok (`int`, *optional*, defaults to 6): Number of experts to route each token to. This is the top-k value for the token-choice routing. num_shared_experts (`int`, *optional*, defaults to 2): Number of shared experts that are always activated for all tokens. route_scale (`float`, *optional*, defaults to 1.0): Scaling factor applied to routing weights. global_attn_every_n_layers (`int`, *optional*, defaults to 4): The frequency of full attention layers. Every Nth layer will use full attention, while others use sliding window attention. sliding_window (`int`, *optional*, defaults to 1024): Sliding window size for local attention layers. layer_types (`list[str]`, *optional*): A list that explicitly maps each layer index with its attention type. Each element should be either "sliding_attention" or "full_attention". If not provided, it will be automatically generated based on `global_attn_every_n_layers`. attention_dropout (`float`, *optional*, defaults to 0.0): The dropout ratio for the attention probabilities. mup_enabled (`bool`, *optional*, defaults to `False`): Whether to enable muP (Maximal Update Parametrization) input scaling. When enabled, input embeddings are scaled by `sqrt(hidden_size)`. eos_token_id (`int`, *optional*): End of stream token id. pad_token_id (`int`, *optional*): Padding token id. bos_token_id (`int`, *optional*): Beginning of stream token id. Example: ```python >>> from transformers import AfmoeModel, AfmoeConfig >>> # Initializing an AFMoE configuration >>> configuration = AfmoeConfig() >>> # Initializing a model from the afmoe-small-sft-v1 style configuration >>> model = AfmoeModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ``` afmoepast_key_values input_ids inputs_embeds hidden_statesattention_mask) embed_tokenslayersnormN vocab_size hidden_sizeintermediate_sizemoe_intermediate_sizenum_hidden_layersnum_dense_layersnum_attention_headsnum_key_value_headshead_dim hidden_actmax_position_embeddingsinitializer_range rms_norm_eps use_cachetie_word_embeddings rope_thetarope_parameters num_expertsnum_experts_per_toknum_shared_experts route_scaleglobal_attn_every_n_layerssliding_window layer_typesattention_dropout mup_enabled eos_token_id pad_token_id bos_token_idc ||_| |_||_||_||_||_||_| |_| |_| |_ | |_ ||_ ||_ ||_ ||_||_||_||_||_d|_||_||_||_||_||_|j0;t3|jDcgc]}t5|dz|zrdndc}|_t7|j0||}||_||_||_||_||_ tC |di|ycc}w)NFsliding_attentionfull_attention)#rrrrrrrrrrrrr!r"rr$r#r%r&attention_biasr*r'r(r+r)rangeboolrrr,r-r.r super__init__)!selfrrrrrrrrrrrrrrr r!r"r#r$r%r&r'r(r)r*r+r,r-r.kwargsi __class__s! _/opt/pipecat/venv/lib/python3.12/site-packages/transformers/models/afmoe/configuration_afmoe.pyr8zAfmoeConfig.__init__srB%'>$&!2!2 0#6   $!2("$.&;"#6 &"4&#"3*D',&&    #t556 (,QU6P,P'Q#Wgg D  d../  &"5 #6 (((#6  "6" sE)iiii r0Nsilui@g{Gz?gh㈵>TFg@N@g?iNgFNNN)__name__ __module__ __qualname____doc__ model_typekeys_to_ignore_at_inferencebase_model_pp_planintstrfloatr6rdictlistr8 __classcell__)r<s@r=rrsvcJJ#4"5&(9:#%568IJ!"_$56"("&(,,0(*'(*,*."!'.3*.%)!%+0#*MQ"$*+)*$'12%)#'*-#($($($(=N#$JN#4ZN#: N# #Tz N# : N#*N#!4ZN#!4ZN#*N#$JN#"%tN#!4<N#dlN#$;N# "D[!N#"DL#N#$($sN/B*CCdJ%N#&4Z'N#(!4Z)N#* $J+N#,T\-N#.%($J/N#0d 1N#2D[3N#4!4<5N#6D[7N#8Tk9N#:Tk;N#<Tk=N#N#rN) rIconfiguration_utilsrrmodeling_rope_utilsrutilsr get_loggerrFloggerr__all__r3rSr=rZs@ J1   H %~#"~#B /rS