qim3ldZddlmZddlmZddlmZejeZ GddeZ dgZ y)z$GraniteMoeHybrid model configuration)PreTrainedConfig)RopeParameters)loggingcUdeZdZdZdZddiZdgZdddd d d d d d ddd dddd dddddddddddd d ddddddddddddedff)ded zded zd ed zd!ed zd"ed zd#ed zd$e d zd%ed zd&ed zd'ed zd(e d zd)ed zd*ed zd+ed zd,e d zd-e e e e fzd zd.e d zd/ed zd0ed zd1ed zd2ed zd3ed zd4ed zd5ed zd6e d zd7ed zd8ed zd9e d zde e d zd:ed zd;ed zded zd?ed zd@ed zdAe d zdBe d zdCed zdDed zdEeeefd zfRfdF ZedGZxZS)HGraniteMoeHybridConfiga: This is the configuration class to store the configuration of a [`GraniteMoeHybridConfig`]. It is used to instantiate an GraniteMoeHybrid model according to the specified arguments, defining the model 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 32000): Vocabulary size of the GraniteMoeHybrid model. Defines the number of different tokens that can be represented by the `inputs_ids` passed when calling [`GraniteMoeHybridModel`] hidden_size (`int`, *optional*, defaults to 4096): Dimension of the hidden representations. intermediate_size (`int`, *optional*, defaults to 11008): Dimension of the MLP representations. num_hidden_layers (`int`, *optional*, defaults to 32): 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*): 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`. hidden_act (`str` or `function`, *optional*, defaults to `"silu"`): The non-linear activation function (function or string) in the decoder. max_position_embeddings (`int`, *optional*, defaults to 2048): 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-06): 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`. pad_token_id (`int`, *optional*): Padding token id. bos_token_id (`int`, *optional*, defaults to 1): Beginning of stream token id. eos_token_id (`int`, *optional*, defaults to 2): End of stream token id. tie_word_embeddings (`bool`, *optional*, defaults to `False`): Whether to tie weight 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`. attention_bias (`bool`, *optional*, defaults to `False`): Whether to use a bias in the query, key, value and output projection layers during self-attention. attention_dropout (`float`, *optional*, defaults to 0.0): The dropout ratio for the attention probabilities. embedding_multiplier (`float`, *optional*, defaults to 1.0): embedding multiplier. logits_scaling (`float`, *optional*, defaults to 1.0): divisor for output logits. residual_multiplier (`float`, *optional*, defaults to 1.0): residual multiplier. attention_multiplier (`float`, *optional*, defaults to 1.0): attention multiplier. num_local_experts (`int`, *optional*, defaults to 8): total number of experts. num_experts_per_tok (`int`, *optional*, defaults to 2): number of experts per token. output_router_logits (`bool`, *optional*, defaults to `False`): Whether or not the router logits should be returned by the model. Enabling this will also allow the model to output the auxiliary loss. router_aux_loss_coef (`float`, *optional*, defaults to 0.001): router auxiliary loss coefficient shared_intermediate_size (`int`, *optional*, defaults to 1024): intermediate size for shared experts. position_embedding_type (`str`, *optional*): Positional embedding type to be used; defaults to None. Allowed options: `[None, "rope"]` layer_types (`List`, *optional*): list of strings to be used as layer types. Allowed choices: "mamba", "attention". mamba_n_heads (`int`, *optional*, defaults to 128): The number of mamba heads used. mamba_n_groups (`int`, *optional*, defaults to 1): The number of the mamba groups used. mamba_d_state (`int`, *optional*, defaults to 256): The dimension the mamba latent state space. mamba_d_head (`int`, *optional*, defaults to `"auto"`): Head embedding dimension size. mamba_d_conv (`int`, *optional*, defaults to 4): The size of the mamba convolution kernel. mamba_expand (`int`, *optional*, defaults to 2): Expanding factor (relative to hidden_size) used to determine the mamba intermediate size. mamba_chunk_size (`int`, *optional*, defaults to 256): The chunks in which to break the sequence when doing prefill/training. mamba_conv_bias (`bool`, *optional*, defaults to `True`): Flag indicating whether or not to use bias in the convolution layer of the mamba mixer block. mamba_proj_bias (`bool`, *optional*, defaults to `False`): Flag indicating whether or not to use bias in the input and output projections (["in_proj", "out_proj"]) of the mamba mixer block. time_step_min (`float`, *optional*, defaults to 0.001): Minimum `time_step` used to bound `dt_proj.bias`. time_step_max (`float`, *optional*, defaults to 0.1): Maximum `time_step` used to bound `dt_proj.bias`. time_step_limit (`tuple`, *optional*, defaults to `(0.0, inf)`): Accepted range of time step values for clamping. ```python >>> from transformers import GraniteMoeHybridModel, GraniteMoeHybridConfig >>> # Initializing a GraniteMoeHybrid config >>> configuration = GraniteMoeHybridConfig() >>> # Accessing the model configuration >>> configuration = model.config ```granitemoehybridlayers_block_type layer_typespast_key_valuesi}ii+ Nsiluig{Gz?gư>TFgg?gMbP?iautog?inf vocab_size hidden_sizeintermediate_sizenum_hidden_layersnum_attention_headsnum_key_value_heads hidden_actmax_position_embeddingsinitializer_range rms_norm_eps use_cache pad_token_id bos_token_id eos_token_idtie_word_embeddingsrope_parametersattention_biasattention_dropoutembedding_multiplierlogits_scalingresidual_multiplierattention_multipliernum_local_expertsnum_experts_per_tokoutput_router_logitsrouter_aux_loss_coefshared_intermediate_sizeposition_embedding_type mamba_n_headsmamba_n_groups mamba_d_state mamba_d_head mamba_d_conv mamba_expandmamba_chunk_sizemamba_conv_biasmamba_proj_bias time_step_min time_step_maxtime_step_limitc* ,||_||_||_||_||_||_||}||_||_| |_| |_ | |_ ||_ ||_ ||_ ||_||_||_||_||_||_||_||_||_||_|#|z}+|t1d|Dr t3d|+|zdk7r t3d|!dk(r|+|z}!|!|z|+k7r t3d||_|!|_||_| |_|"|_|$|_|%|_ |&|_!|'|_"|(|_#|) tI|)nd|_%|#|_&||_'||_(| |_)| |_*||_+tY,|di|*y)Nc3$K|]}|dv yw))mamba attentionN).0 layer_types u/opt/pipecat/venv/lib/python3.12/site-packages/transformers/models/granitemoehybrid/configuration_granitemoehybrid.py z2GraniteMoeHybridConfig.__init__..s*rXb:=S+S*rsz$&!2!2#6   &"5 #6 $!2(",$8!,#6 $8!!2!2#6 $8!$8!(@%'>$.)K7  "s*rfq*r'r[\ \  - 2ST T 6 !->L - '+= =op p*(,*( 0..**9H9Tu_5Z^(&#6 ((( "6"cR|jr |jSdg|jzS)Nr@)r r)rMs rEr z(GraniteMoeHybridConfig.layers_block_types(#'#3#3t['TE[E[9[[rQ)__name__ __module__ __qualname____doc__ model_type attribute_mapkeys_to_ignore_at_inferencefloatintstrboolrdictlistrJrLpropertyr __classcell__)rPs@rErrsfh T$J]M$5"5"'"&(-(**,*.!'.2*.#'!%#'#$#$+0MQ&+*--0'*,/-0()*+,1-2/3.2(,$'%&$'#)#$#$'*'+',&+&)7:E%L6IUl#$Jl#4Zl#: l# : l# !4Z l#!4Zl#$Jl#"%tl#!4<l#Djl#$;l#Djl#Djl#Djl# "D[!l#"($sN/B*CCdJ#l#$t %l#&!4<'l#($dl)l#* +l#,#T\-l#.$dl/l#0:1l#2!4Z3l#4#Tk5l#6$dl7l#8#&*9l#:"%t;l#<#Y%=l#>Tz?l#@d Al#BTzCl#DDjEl#FDjGl#HDjIl#J*Kl#LMl#NOl#Pt|Ql#Rt|Sl#Tue|,t3Ul#^\\rQrN) rVconfiguration_utilsrmodeling_rope_utilsrutilsr get_loggerrSloggerr__all__rBrQrErhsC+31   H %b\-b\J $ $rQ