o ZhJ@sdZddlmZddlmZmZmZmZddlm Z ddl m Z ddl m Z er4ddlmZdd l mZe eZGd d d e ZGd d d e ZGddde ZGddde ZgdZdS)zGroupViT model configuration OrderedDict) TYPE_CHECKINGAnyMappingOptional)PretrainedConfig) OnnxConfig)logging)ProcessorMixin) TensorTypecsHeZdZdZdZdZ         dfdd ZZS)GroupViTTextConfiga> This is the configuration class to store the configuration of a [`GroupViTTextModel`]. It is used to instantiate an GroupViT 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 GroupViT [nvidia/groupvit-gcc-yfcc](https://huggingface.co/nvidia/groupvit-gcc-yfcc) 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 49408): Vocabulary size of the GroupViT text model. Defines the number of different tokens that can be represented by the `inputs_ids` passed when calling [`GroupViTModel`]. hidden_size (`int`, *optional*, defaults to 256): Dimensionality of the encoder layers and the pooler layer. intermediate_size (`int`, *optional*, defaults to 1024): Dimensionality of the "intermediate" (i.e., feed-forward) layer in the Transformer encoder. num_hidden_layers (`int`, *optional*, defaults to 12): Number of hidden layers in the Transformer encoder. num_attention_heads (`int`, *optional*, defaults to 4): Number of attention heads for each attention layer in the Transformer encoder. max_position_embeddings (`int`, *optional*, defaults to 77): 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). hidden_act (`str` or `function`, *optional*, defaults to `"quick_gelu"`): The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`, `"relu"`, `"selu"` and `"gelu_new"` `"quick_gelu"` are supported. layer_norm_eps (`float`, *optional*, defaults to 1e-5): The epsilon used by the layer normalization layers. attention_dropout (`float`, *optional*, defaults to 0.0): The dropout ratio for the attention probabilities. dropout (`float`, *optional*, defaults to 0.0): The dropout probability for all fully connected layers in the embeddings, encoder, and pooler. initializer_range (`float`, *optional*, defaults to 0.02): The standard deviation of the truncated_normal_initializer for initializing all weight matrices. initializer_factor (`float`, *optional*, defaults to 1.0): A factor for initializing all weight matrices (should be kept to 1, used internally for initialization testing). Example: ```python >>> from transformers import GroupViTTextConfig, GroupViTTextModel >>> # Initializing a GroupViTTextModel with nvidia/groupvit-gcc-yfcc style configuration >>> configuration = GroupViTTextConfig() >>> model = GroupViTTextModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ```Zgroupvit_text_model text_config M quick_geluh㈵>{Gz??c sftjd| ||d|||_||_||_| |_||_||_||_||_ ||_ | |_ | |_ | |_ dS)N) pad_token_id bos_token_id eos_token_id)super__init__ vocab_size hidden_sizeintermediate_sizedropoutnum_hidden_layersnum_attention_headsmax_position_embeddingslayer_norm_eps hidden_actinitializer_rangeinitializer_factorattention_dropout)selfr$r%r&r(r)r*r,r+r'r/r-r.rrr kwargs __class__r!b/var/www/auris/lib/python3.10/site-packages/transformers/models/groupvit/configuration_groupvit.pyr#Zs zGroupViTTextConfig.__init__)rrrrrrrrrrrrrrr__name__ __module__ __qualname____doc__ model_typeZbase_config_keyr# __classcell__r!r!r2r4r!s(5rcs^eZdZdZdZdZddgddgdgd d d d d dddddddddgffdd ZZS)GroupViTVisionConfiga@ This is the configuration class to store the configuration of a [`GroupViTVisionModel`]. It is used to instantiate an GroupViT 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 GroupViT [nvidia/groupvit-gcc-yfcc](https://huggingface.co/nvidia/groupvit-gcc-yfcc) architecture. Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the documentation from [`PretrainedConfig`] for more information. Args: hidden_size (`int`, *optional*, defaults to 384): Dimensionality of the encoder layers and the pooler layer. intermediate_size (`int`, *optional*, defaults to 1536): Dimensionality of the "intermediate" (i.e., feed-forward) layer in the Transformer encoder. depths (`List[int]`, *optional*, defaults to [6, 3, 3]): The number of layers in each encoder block. num_group_tokens (`List[int]`, *optional*, defaults to [64, 8, 0]): The number of group tokens for each stage. num_output_groups (`List[int]`, *optional*, defaults to [64, 8, 8]): The number of output groups for each stage, 0 means no group. num_attention_heads (`int`, *optional*, defaults to 6): Number of attention heads for each attention layer in the Transformer encoder. image_size (`int`, *optional*, defaults to 224): The size (resolution) of each image. patch_size (`int`, *optional*, defaults to 16): The size (resolution) of each patch. hidden_act (`str` or `function`, *optional*, defaults to `"gelu"`): The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`, `"relu"`, `"selu"` and `"gelu_new"` `"quick_gelu"` are supported. layer_norm_eps (`float`, *optional*, defaults to 1e-5): The epsilon used by the layer normalization layers. dropout (`float`, *optional*, defaults to 0.0): The dropout probability for all fully connected layers in the embeddings, encoder, and pooler. attention_dropout (`float`, *optional*, defaults to 0.0): The dropout ratio for the attention probabilities. initializer_range (`float`, *optional*, defaults to 0.02): The standard deviation of the truncated_normal_initializer for initializing all weight matrices. initializer_factor (`float`, *optional*, defaults to 1.0): A factor for initializing all weight matrices (should be kept to 1, used internally for initialization testing). Example: ```python >>> from transformers import GroupViTVisionConfig, GroupViTVisionModel >>> # Initializing a GroupViTVisionModel with nvidia/groupvit-gcc-yfcc style configuration >>> configuration = GroupViTVisionConfig() >>> model = GroupViTVisionModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ```Zgroupvit_vision_model vision_configii)rrr)@r)r?r@r@r>rZgelurrrrg?rc stjdi|||_||_||_|t|kr%td|dt|||_||_ ||_ ||_ ||_ | |_ | |_| |_| |_| |_||_||_||_||_||_dS)Nz&Manually setting num_hidden_layers to z1, but we expect num_hidden_layers = sum(depth) = r!)r"r#r%r&depthssumloggerwarningr(num_group_tokensnum_output_groupsr) image_size patch_size num_channelsr,r+r'r/r-r. assign_epsassign_mlp_ratio)r0r%r&rCr(rGrHr)rIrJrKr,r+r'r/r-r.rLrMr1r2r!r4r#s2  zGroupViTVisionConfig.__init__r5r!r!r2r4r<}s.7r<csPeZdZdZdZeedZ     dfdd Ze d ed efd d Z Z S)GroupViTConfiga [`GroupViTConfig`] is the configuration class to store the configuration of a [`GroupViTModel`]. It is used to instantiate a GroupViT model according to the specified arguments, defining the text model and vision model configs. Instantiating a configuration with the defaults will yield a similar configuration to that of the GroupViT [nvidia/groupvit-gcc-yfcc](https://huggingface.co/nvidia/groupvit-gcc-yfcc) architecture. Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the documentation from [`PretrainedConfig`] for more information. Args: text_config (`dict`, *optional*): Dictionary of configuration options used to initialize [`GroupViTTextConfig`]. vision_config (`dict`, *optional*): Dictionary of configuration options used to initialize [`GroupViTVisionConfig`]. projection_dim (`int`, *optional*, defaults to 256): Dimensionality of text and vision projection layers. projection_intermediate_dim (`int`, *optional*, defaults to 4096): Dimensionality of intermediate layer of text and vision projection layers. logit_scale_init_value (`float`, *optional*, defaults to 2.6592): The initial value of the *logit_scale* parameter. Default is used as per the original GroupViT implementation. kwargs (*optional*): Dictionary of keyword arguments. Zgroupvitrr=Nr/L F@c s|dd}|dd}tjdi||dur]|duri}tdi|} | D]+\} } | |vrW| || krW| dvrW| |vrLd| d| d} nd| d} t| q,|| |dur|durgi}t di|} d | vrd d | d D| d <| D]+\} } | |vr| || kr| dvr| |vrd| d | d} nd | d} t| q|| |duri}td|duri}tdtdi||_ t di||_ ||_ ||_ ||_d|_d|_d|_dS)Ntext_config_dictvision_config_dict)Ztransformers_version`zp` is found in both `text_config_dict` and `text_config` but with different values. The value `text_config_dict["z"]` will be used instead.zn`text_config_dict` is provided which will be used to initialize `GroupViTTextConfig`. The value `text_config["z"]` will be overridden.Zid2labelcSsi|] \}}t||qSr!)str).0keyvaluer!r!r4 =sz+GroupViTConfig.__init__..zv` is found in both `vision_config_dict` and `vision_config` but with different values. The value `vision_config_dict["zt`vision_config_dict` is provided which will be used to initialize `GroupViTVisionConfig`. The value `vision_config["zS`text_config` is `None`. Initializing the `GroupViTTextConfig` with default values.zW`vision_config` is `None`. initializing the `GroupViTVisionConfig` with default values.rrFr!)popr"r#rto_dictitemsrEinfoupdater<rr=projection_dimprojection_intermediate_dimlogit_scale_init_valuer-r.Zoutput_segmentation)r0rr=r_r`rar1rRrSZ_text_config_dictrWrXmessageZ_vision_config_dictr2r!r4r#sr          zGroupViTConfig.__init__rr=cKs|d||d|S)z Instantiate a [`GroupViTConfig`] (or a derived class) from groupvit text model configuration and groupvit vision model configuration. Returns: [`GroupViTConfig`]: An instance of a configuration object rONr!)r[)clsrr=r1r!r!r4from_text_vision_configsgs z'GroupViTConfig.from_text_vision_configs)NNrrPrQ) r6r7r8r9r:rr<Z sub_configsr# classmethodrdr;r!r!r2r4rNs arNc seZdZedeeeeefffddZedeeeeefffddZede fddZ   dd d d ed ede ddeee ff fdd Z edefddZZS)GroupViTOnnxConfigreturncCs0tddddfdddddd fd dddfgS) NZ input_idsbatchsequence)rrZ pixel_valuesrKheightwidth)rrrZattention_maskrr0r!r!r4inputsus   zGroupViTOnnxConfig.inputscCs0tdddifdddifdddifdddifgS)NZlogits_per_imagerrhZlogits_per_textZ text_embedsZ image_embedsrrmr!r!r4outputss    zGroupViTOnnxConfig.outputscCdS)Ng-C6?r!rmr!r!r4atol_for_validationz&GroupViTOnnxConfig.atol_for_validationN processorr batch_size seq_length frameworkr cs6tj|j|||d}tj|j||d}i||S)N)rurvrw)rurw)r"generate_dummy_inputsZ tokenizerZimage_processor)r0rtrurvrwZtext_input_dictZimage_input_dictr2r!r4rxs  z(GroupViTOnnxConfig.generate_dummy_inputscCrp)Nr!rmr!r!r4default_onnx_opsetrrz%GroupViTOnnxConfig.default_onnx_opset)rsrsN)r6r7r8propertyrrUintrnrofloatrqrrrxrzr;r!r!r2r4rfts.    rf)rNrfrr<N)r9 collectionsrtypingrrrrZconfiguration_utilsr Zonnxr utilsr Zprocessing_utilsr r Z get_loggerr6rErr<rNrf__all__r!r!r!r4s        \l .