qik-dZddlmZddlmZmZddlmZmZeje Z dZ GddeZ ee jd d d Gd d e Zee jdd dGdde Zee jdd dGdde ZGddeZgdZy)zBARK model configuration)PreTrainedConfig)add_start_docstringslogging)CONFIG_MAPPING AutoConfiga This is the configuration class to store the configuration of a [`{model}`]. It is used to instantiate the 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 Bark [suno/bark](https://huggingface.co/suno/bark) architecture. Configuration objects inherit from [`PreTrainedConfig`] and can be used to control the model outputs. Read the documentation from [`PreTrainedConfig`] for more information. Args: block_size (`int`, *optional*, defaults to 1024): 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). input_vocab_size (`int`, *optional*, defaults to 10_048): Vocabulary size of a Bark sub-model. Defines the number of different tokens that can be represented by the `inputs_ids` passed when calling [`{model}`]. Defaults to 10_048 but should be carefully thought with regards to the chosen sub-model. output_vocab_size (`int`, *optional*, defaults to 10_048): Output vocabulary size of a Bark sub-model. Defines the number of different tokens that can be represented by the: `output_ids` when passing forward a [`{model}`]. Defaults to 10_048 but should be carefully thought with regards to the chosen sub-model. num_layers (`int`, *optional*, defaults to 12): Number of hidden layers in the given sub-model. num_heads (`int`, *optional*, defaults to 12): Number of attention heads for each attention layer in the Transformer architecture. hidden_size (`int`, *optional*, defaults to 768): Dimensionality of the "intermediate" (often named feed-forward) layer in the architecture. dropout (`float`, *optional*, defaults to 0.0): The dropout probability for all fully connected layers in the embeddings, encoder, and pooler. bias (`bool`, *optional*, defaults to `True`): Whether or not to use bias in the linear layers and layer norm layers. initializer_range (`float`, *optional*, defaults to 0.02): The standard deviation of the truncated_normal_initializer for initializing all weight matrices. use_cache (`bool`, *optional*, defaults to `True`): Whether or not the model should return the last key/values attentions (not used by all models). cHeZdZdgZdddddZ dfd ZxZS) BarkSubModelConfigpast_key_values num_heads num_layersinput_vocab_size block_size)num_attention_headsnum_hidden_layers vocab_size window_sizec  ||_||_||_||_||_||_||_||_| |_| |_ t |,di| yN) rroutput_vocab_sizer r hidden_sizedropoutbias use_cacheinitializer_rangesuper__init__) selfrrrr r rrrrrkwargs __class__s ]/opt/pipecat/venv/lib/python3.12/site-packages/transformers/models/bark/configuration_bark.pyrzBarkSubModelConfig.__init__Hs_% 0!2$"&  "!2 "6") i@'r$ r%igT{Gz?T)__name__ __module__ __qualname__keys_to_ignore_at_inference attribute_mapr __classcell__r!s@r"r r >sK#4"5 +)(# M  ##r#r BarkSemanticConfigBarkSemanticModel)configmodela Example: ```python >>> from transformers import BarkSemanticConfig, BarkSemanticModel >>> # Initializing a Bark sub-module style configuration >>> configuration = BarkSemanticConfig() >>> # Initializing a model (with random weights) from the suno/bark style configuration >>> model = BarkSemanticModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ```ceZdZdZdZy)r.semanticsemantic_configNr'r(r) model_typebase_config_keyrr#r"r.r.ds&J'Or#BarkCoarseConfigBarkCoarseModela Example: ```python >>> from transformers import BarkCoarseConfig, BarkCoarseModel >>> # Initializing a Bark sub-module style configuration >>> configuration = BarkCoarseConfig() >>> # Initializing a model (with random weights) from the suno/bark style configuration >>> model = BarkCoarseModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ```ceZdZdZdZy)r8coarse_acousticscoarse_acoustics_configNr5rr#r"r8r8{s&$J/Or#BarkFineConfig BarkFineModela n_codes_total (`int`, *optional*, defaults to 8): The total number of audio codebooks predicted. Used in the fine acoustics sub-model. n_codes_given (`int`, *optional*, defaults to 1): The number of audio codebooks predicted in the coarse acoustics sub-model. Used in the acoustics sub-models. Example: ```python >>> from transformers import BarkFineConfig, BarkFineModel >>> # Initializing a Bark sub-module style configuration >>> configuration = BarkFineConfig() >>> # Initializing a model (with random weights) from the suno/bark style configuration >>> model = BarkFineModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ```c(eZdZdZdZdfd ZxZS)r=fine_acousticsfine_acoustics_configc N||_||_||_t|di|yr) n_codes_total n_codes_giventie_word_embeddingsrr)rrErCrDr r!s r"rzBarkFineConfig.__init__s+**#6  "6"r#)T)r'r(r)r6r7rr,r-s@r"r=r=s0"J-O##r#c jeZdZdZdZeeeedZ d de dzde dzde dzde dzffd Z xZ S) BarkConfiga This is the configuration class to store the configuration of a [`BarkModel`]. It is used to instantiate a Bark model according to the specified sub-models configurations, defining the model architecture. Instantiating a configuration with the defaults will yield a similar configuration to that of the Bark [suno/bark](https://huggingface.co/suno/bark) architecture. Configuration objects inherit from [`PreTrainedConfig`] and can be used to control the model outputs. Read the documentation from [`PreTrainedConfig`] for more information. Args: semantic_config ([`BarkSemanticConfig`], *optional*): Configuration of the underlying semantic sub-model. coarse_acoustics_config ([`BarkCoarseConfig`], *optional*): Configuration of the underlying coarse acoustics sub-model. fine_acoustics_config ([`BarkFineConfig`], *optional*): Configuration of the underlying fine acoustics sub-model. codec_config ([`AutoConfig`], *optional*): Configuration of the underlying codec sub-model. Example: ```python >>> from transformers import ( ... BarkSemanticConfig, ... BarkCoarseConfig, ... BarkFineConfig, ... BarkModel, ... BarkConfig, ... AutoConfig, ... ) >>> # Initializing Bark sub-modules configurations. >>> semantic_config = BarkSemanticConfig() >>> coarse_acoustics_config = BarkCoarseConfig() >>> fine_acoustics_config = BarkFineConfig() >>> codec_config = AutoConfig.from_pretrained("facebook/encodec_24khz") >>> # Initializing a Bark module style configuration >>> configuration = BarkConfig( ... semantic_config, coarse_acoustics_config, fine_acoustics_config, codec_config ... ) >>> # Initializing a model (with random weights) >>> model = BarkModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ``` bark)r4r<rA codec_configNr4r<rArKc | t}tjdnt|tr tdi|}| t }tjdnt|tr t di|}| t }tjdnt|tr t di|}|#td}tjdn0t|tr |jdd}t|di|}||_ ||_ ||_ ||_ ||_ t|<di|y)NzW`semantic_config` is `None`. Initializing the `BarkSemanticConfig` with default values.z]`coarse_acoustics_config` is `None`. Initializing the `BarkCoarseConfig` with default values.zY`fine_acoustics_config` is `None`. Initializing the `BarkFineConfig` with default values.encodeczN`codec_config` is `None`. Initializing the `codec_config` with default values.r6r)r.loggerinfo isinstancedictr8r=rgetr4r<rArKrrr) rr4r<rArKrr codec_model_typer!s r"rzBarkConfig.__init__s/  "02O KKq r  .0C?CO " *&6&8 # KKo / 6&6&Q9P&Q # ($2$4 ! KKs t -t 4$2$K5J$K !  ))46L KKh i  d ++// iH )*:;KlKL.'>$%:"(!2 "6"r#)NNNNr&) r'r(r)__doc__r6r.r8r=r sub_configsrQrr,r-s@r"rIrIst2hJ-#3!/" K(,/3-1$( +#+#"&+# $d{ +# Tk +#+#r#rI)r8rIr=r.N)rTconfiguration_utilsrutilsrrautorr get_loggerr'rN#BARK_SUBMODELCONFIG_START_DOCSTRINGr formatr.r8r=rI__all__rr#r"r]s32-   H %#'#L##)##L'..6JRe.f $(+(%$( '..6HPa.b $0)0%$0 '..6Fo.^ . #' #/. #h#!h#V Ur#