^jH) dZddlZddlmZddlmZddlmZmZeje Z edeGd d eZ edeGd d eZ edeGd deZgdZy)zCLVP model configurationN)strict)PreTrainedConfig)auto_docstringloggingzsusnato/clvp_dev) checkpointcneZdZUdZdZddgZdZeed<dZ eed<d Z eed <dZ eed <d Z eed <dZ eed<dZeed<dZeed<dZeezed<dZeezed<dZeed<dZeed<dZeed<dZeed<dZed zed!<d"Zeeezd zed#<d Zed zed$<e d(d%eej>zd&efd'Z y ))ClvpEncoderConfiga use_rotary_embedding (`bool`, *optional*, defaults to `True`): Whether to use rotary_embedding or not. use_attention_bias (`bool`, *optional*, defaults to `False`): Whether to use bias in Query, Key and Value layers during self attention. summary_type (`str`, *optional*, defaults to `"mean"`): What strategy to use to get pooler_output from the last_hidden_state. `"last"`, `"first"`, `"mean"` and `"cls_index"` are supported. Example: ```python >>> from transformers import ClvpEncoderConfig, ClvpEncoder >>> # Initializing a ClvpEncoderConfig with susnato/clvp_dev style configuration >>> encoder_configuration = ClvpEncoderConfig() >>> # Initializing a ClvpEncoder (with random weights) from the susnato/clvp_dev style configuration >>> model = ClvpEncoder(encoder_configuration) >>> # Accessing the model configuration >>> configuration = model.config ``` clvp_encoder text_config speech_config vocab_size hidden_sizeiintermediate_sizeprojection_dimnum_hidden_layers num_attention_headsgelu hidden_acth㈵>layer_norm_eps皙?attention_dropoutdropoutTuse_rotary_embeddingFuse_attention_biasmean summary_type?initializer_factorN bos_token_idr eos_token_id pad_token_idpretrained_model_name_or_path config_typec T|j|fi|\}}||jvrtd||jddk(r||}d|vrGt |dr;|d|j k7r)t jd|dd|j d|j|fi|S)NzSWe can only load either 'text_config' or 'speech_config' but you are trying to load model_typeclvpzYou are using a model of type z to instantiate a model of type zN. This is not supported for all configurations of models and can yield errors.) get_config_dictbase_config_key ValueErrorgethasattrr,loggerwarning from_dict)clsr)r*kwargs config_dicts v/var/www/ramen.bs-engineer-server.com/venv/lib/python3.12/site-packages/transformers/models/clvp/configuration_clvp.pyfrom_pretrainedz!ClvpEncoderConfig.from_pretrainedKs2c112OZSYZ V c11 1efqers  ??< (F 2%k2K ; &73 +E+VbJcgjguguJu NN0\1J0KKk>>""pr  s}}[3F33)r )!__name__ __module__ __qualname____doc__r,r/rint__annotations__rrrrrrstrrfloatrrrboolr r"r$r&r'listr( classmethodosPathLiker:r;r9r r s 0 J$o6OJK!s!NCs!!J NE %(us{(GUS[!%$%$$L# ##"L#*"+,L#S /D(,#L#*#R_4,/"++,=4LO44r;r ceZdZUdZdZdZdZeed<dZ eed<dZ eed <d Z eed <d Z eed <dZ eed<dZedzed<dZeed<dZeed<dZeezed<dZeezed<dZeezed<dZeed<dZeed<dZeed<d Zeed!<dZedzed"<d Zeed#<dZeezed$<d Zeed%<d&Zedzed'<d(Z ee!ezdzed)<dZ"edzed*<d+Z#eed,<d Z$eed-<d.Z%eed/<d0Z&e!ee'ed1fzed2<d3Z(eed4<y)5ClvpDecoderConfiga% max_text_tokens (`int`, *optional*, defaults to 404): The maximum sequence length of text tokens that this model might ever be used with. Similar to `n_positions` in `GPT2Config`. n_inner (`int`, *optional*): Dimensionality of the inner feed-forward layers. `None` will set it to 4 times `hidden_size`. num_mel_attn_blocks (`int`, *optional*, defaults to 6): Denotes the number of self attention layers in [`ClvpConditioningEncoder`]. summary_type (`string`, *optional*, defaults to `"cls_index"`): Argument used when doing sequence summary. Has to be one of the following options: - `"last"`: Take the last token hidden state (like XLNet). - `"first"`: Take the first token hidden state (like BERT). - `"mean"`: Take the mean of all tokens hidden states. - `"cls_index"`: Supply a Tensor of classification token position (like GPT/GPT-2). - `"attn"`: Not implemented now, use multi-head attention. summary_use_proj (`bool`, *optional*, defaults to `True`): Whether or not to add a projection after the vector extraction. summary_activation (`str`, *optional*): Pass `"tanh"` for a tanh activation to the output, any other value will result in no activation. summary_proj_to_labels (`bool`, *optional*, defaults to `True`): Whether the projection outputs should have `config.num_labels` or `config.hidden_size` classes. summary_first_dropout (`float`, *optional*, defaults to 0.1): The dropout ratio to be used after the projection and activation. feature_size (`int`, *optional*, defaults to 80): The feature dimension of the extracted mel features. This value is used in [`ClvpConditioningEncoder`]. use_attention_bias (`bool`, *optional*, defaults to `True`): Whether to use bias in Query, Key and Value layers during self attention. decoder_fixing_codes (`list`, *optional*, defaults to `[83, 45, 45, 248]`): These values are used in the method `fix_speech_decoder_output` to fix decoder generated outputs. Example: ```python >>> from transformers import ClvpDecoderConfig, ClvpDecoder >>> # Initializing a ClvpDecoderConfig with susnato/clvp_dev style configuration >>> decoder_configuration = ClvpDecoderConfig() >>> # Initializing a ClvpDecoder (with random weights) from the susnato/clvp_dev style configuration >>> model = ClvpDecoder(decoder_configuration) >>> # Accessing the model configuration >>> configuration = model.config ``` clvp_decoderdecoder_configi ri`max_position_embeddingsimax_text_tokensirrrNn_innernum_mel_attn_blocksgelu_newactivation_functionr resid_pdrop embd_pdroprrlayer_norm_epsilong{Gz?initializer_range cls_indexr"Tsummary_use_projsummary_activationsummary_proj_to_labelssummary_first_dropout use_cachei r&i r'r(P feature_sizer r#r$)S-rd.decoder_fixing_codesFadd_cross_attention))r<r=r>r?r,r/rr@rArNrOrrrrRrTrVrBrWrCrXrrYrZr"r\rDr]r^r_r`r&r'rEr(rbr r$rftuplergrIr;r9rKrKesz, \ J&OJ#&S&OSKs!!GS4Z  ))"K"!J !%(us{( $$#u##L##!d!%)d )#'D'),53;,It#L#*#+/L#S /D(/#L#*#L### ##8I$s)eCHo5I %%r;rKceZdZUdZdZeeedZdZe e zdze d<dZ e e zdze d<dZ e e zdze d<dZee d <d Zee d <d Zee d <fdZxZS) ClvpConfiga speech_config (`dict`, *optional*): Dictionary of configuration options used to initialize CLVP speech encoder. decoder_config (`dict`, *optional*): Dictionary of configuration options used to initialize [`ClvpDecoderConfig`]. Example: ```python >>> from transformers import ClvpConfig, ClvpModelForConditionalGeneration >>> # Initializing a ClvpConfig with susnato/clvp_dev style configuration >>> configuration = ClvpConfig() >>> # Initializing a ClvpModelForConditionalGeneration (with random weights) from the susnato/clvp_dev style configuration >>> model = ClvpModelForConditionalGeneration(configuration) >>> # Accessing the model configuration >>> configuration = model.config >>> # We can also initialize a CLVPConfig from a CLVPTextConfig, CLVPSpeechConfig and a CLVPAutoRegressiveConfig >>> from transformers import ClvpEncoderConfig, ClvpDecoderConfig >>> # Initializing a CLVP text, CLVP speech and CLVP decoder configuration >>> config_text = ClvpEncoderConfig() >>> config_speech = ClvpEncoderConfig() >>> decoder_config = ClvpDecoderConfig() >>> config = ClvpConfig(config_text, config_speech, decoder_config) ```r-)r r rMNr r rMrrg/L F@logit_scale_init_valuer#r$c |j%t|_tjdn4t |jt rtdi|j|_|j %t|_tjdn4t |j t rtdi|j |_|j%t|_tjdn4t |jt rtdi|j|_t|(di|y)NzR`text_config` is `None`. initializing the `ClvpEncoderConfig` with default values.zT`speech_config` is `None`. initializing the `ClvpEncoderConfig` with default values.zS`image_config` is `None`. initializing the `ClvpDecoderConfig` with default values.rI) r r r3info isinstancedictr rMrKsuper __post_init__)selfr7 __class__s r9rqzClvpConfig.__post_init__s    #02D  KKl m (($ /0D43C3CDD     %!2!4D  KKn o **D 1!2!HT5G5G!HD     &"3"5D  KKm n ++T 2"3"Jd6I6I"JD  ''r;)r<r=r>r?r,r rK sub_configsr rorrAr rMrr@rkrCr$rq __classcell__)rss@r9rjrjs >J(*+K 37K((4/648M4**T1859ND++d29NC$*E* ##((r;rj)rjrKr )r?rGhuggingface_hub.dataclassesrconfiguration_utilsrutilsrr get_loggerr<r3r rKrj__all__rIr;r9r{s .3,   H %-.E4(E4/E4P-.M&(M&/M&`-.A(!A(/A(H Cr;