IgPdZddlZddlmZddlmZmZmZmZm Z ddl m Z ddl m Z ddlmZer dd lmZdd lmZej&eZGd d e ZGd de ZGdde ZGdde Zy)zGroupViT model configurationN OrderedDict) TYPE_CHECKINGAnyMappingOptionalUnion)PretrainedConfig) OnnxConfig)logging)ProcessorMixin) TensorTypec~eZdZdZdZ dfd Zedeee jfddfdZ xZ S) 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 ```groupvit_text_modelc t|d| ||d|||_||_||_| |_||_||_||_||_ ||_ | |_ | |_ | |_ y)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)selfrrrrrr r"r!rr%r#r$rrrkwargs __class__s p/var/www/html/answerous/venv/lib/python3.12/site-packages/transformers/models/groupvit/configuration_groupvit.pyrzGroupViTTextConfig.__init__Zs}& sl\hslrs$&!2 !2#6 '>$,$!2"4!2pretrained_model_name_or_pathreturnr c >|j||j|fi|\}}|jddk(r|d}d|vrGt|dr;|d|jk7r)t j d|dd|jd|j|fi|S)N model_typegroupvit text_configYou are using a model of type to instantiate a model of type N. This is not supported for all configurations of models and can yield errors._set_token_in_kwargsget_config_dictgethasattrr.loggerwarning from_dictclsr+r' config_dicts r)from_pretrainedz"GroupViTTextConfig.from_pretrained|s   (1c112OZSYZ V ??< (J 6%m4K ; &73 +E+VbJcgjguguJu NN0\1J0KKk>>""pr  s}}[3F33r*)ii M quick_geluh㈵>rF{Gz??ii __name__ __module__ __qualname____doc__r.r classmethodr strosPathLiker? __classcell__r(s@r)rr"sv3 j'J "! 3D4E#r{{BR>> 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 ```groupvit_vision_modelii)r r rA)@r)rYrZrZrXr gelurErFrGrHg?rBc t|di|||_||_||_|t |k7r$t jd|dt |||_||_ ||_ ||_ ||_ | |_ | |_| |_| |_| |_||_||_||_||_||_y)Nz&Manually setting num_hidden_layers to z1, but we expect num_hidden_layers = sum(depth) = r)rrrrdepthssumr9r:rnum_group_tokensnum_output_groupsr image_size patch_size num_channelsr"r!rr%r#r$ assign_epsassign_mlp_ratio)r&rrr_rrarbrrcrdrer"r!rr%r#r$rfrgr'r(s r)rzGroupViTVisionConfig.__init__s, "6"&!2 F + NN89J8KL!!$V / "3 0!2#6 $$($, !2!2"4$ 0r*r+r,r c >|j||j|fi|\}}|jddk(r|d}d|vrGt|dr;|d|jk7r)t j d|dd|jd|j|fi|S)Nr.r/ vision_configr1r2r3r4r<s r)r?z$GroupViTVisionConfig.from_pretraineds   (1c112OZSYZ V ??< (J 6%o6K ; &73 +E+VbJcgjguguJu NN0\1J0KKk>>""pr  s}}[3F33r*rJrTs@r)rVrVs5 n)J#$q'.1`4E#r{{BRUW KK( )"  !4 5  K KKm n  M KKq r-< <1BMB,+F(&<#!%"%#( Q3s2G1r0ric P|d|j|jd|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 )r0rir)rs)r=r0rir's r)from_text_vision_configsz'GroupViTConfig.from_text_vision_configss,f{224MDYDYD[f_effr*)NNr@ig/L F@) rKrLrMrNr.rrOrrVrrSrTs@r)rkrk sO2J$(% _)B g3E gVj g gr*rkc eZdZedeeeeefffdZedeeeeefffdZede fdZ d dddeded e d deee ff fd Z edefd ZxZS)GroupViTOnnxConfigr,c @tddddfdddddd fd dddfgS) N input_idsbatchsequence)rrI pixel_valuesreheightwidth)rrIr attention_maskrr&s r)inputszGroupViTOnnxConfig.inputss@'j9:WHQX!YZ!w:#>?   r*c@tdddifdddifdddifdddifgS)Nlogits_per_imagerrlogits_per_text text_embeds image_embedsrrs r)outputszGroupViTOnnxConfig.outputssD#a\2"QL1G -!W.    r*cy)Ng-C6?rrs r)atol_for_validationz&GroupViTOnnxConfig.atol_for_validationsr* processorr batch_size seq_length frameworkrct||j|||}t||j||}i||S)N)rrr)rr)rgenerate_dummy_inputs tokenizerimage_processor)r&rrrrtext_input_dictimage_input_dictr(s r)rz(GroupViTOnnxConfig.generate_dummy_inputss` '7   J:Yb8 !78  % %* 9 7/6%566r*cy)Nrrs r)default_onnx_opsetz%GroupViTOnnxConfig.default_onnx_opsetsr*)rN)rKrLrMpropertyrrPintrrfloatrrrrrrSrTs@r)rrs  WS#X%6 67   gc3h&7!78  U ,0 7# 7 7 7 L) 7 c  7Cr*r)rNrQ collectionsrtypingrrrrr configuration_utilsr onnxr utilsr processing_utilsrr get_loggerrKr9rrVrkrrr*r)rs|# #??32#   H %j4)j4Zz4+z4zGg%GgT++r*