+ jORt^RIHt^RIHt^RIHt^RIHt^RI H t ]!RR7]!RR ]444t R .t R #) zDPT model configuration)strict)%consolidate_backbone_kwargs_to_config)PreTrainedConfig)auto_docstring) AutoConfigzIntel/dpt-large) checkpointcaa]tRt^toRtRtR]/tRt^ t ^ t Rt Rt Rt RtRtR tR t^t^tR tR tRtR tRtRtRtRtR tRtR tR tRt ^t!Rt"Rt#Rt$Rt%Rt&Rt'V3Rlt(V3Rlt)Rt*Vt+V;t,#) DPTConfiga8 is_hybrid (`bool`, *optional*, defaults to `False`): Whether to use a hybrid backbone. Useful in the context of loading DPT-Hybrid models. backbone_out_indices (`list[int]`, *optional*, defaults to `[2, 5, 8, 11]`): Indices of the intermediate hidden states to use from backbone. readout_type (`str`, *optional*, defaults to `"project"`): The readout type to use when processing the readout token (CLS token) of the intermediate hidden states of the ViT backbone. Can be one of [`"ignore"`, `"add"`, `"project"`]. - "ignore" simply ignores the CLS token. - "add" passes the information from the CLS token to all other tokens by adding the representations. - "project" passes information to the other tokens by concatenating the readout to all other tokens before projecting the representation to the original feature dimension D using a linear layer followed by a GELU non-linearity. reassemble_factors (`list[int]`, *optional*, defaults to `[4, 2, 1, 0.5]`): The up/downsampling factors of the reassemble layers. neck_hidden_sizes (`list[str]`, *optional*, defaults to `[96, 192, 384, 768]`): The hidden sizes to project to for the feature maps of the backbone. fusion_hidden_size (`int`, *optional*, defaults to 256): The number of channels before fusion. head_in_index (`int`, *optional*, defaults to -1): The index of the features to use in the heads. use_batch_norm_in_fusion_residual (`bool`, *optional*, defaults to `False`): Whether to use batch normalization in the pre-activate residual units of the fusion blocks. use_bias_in_fusion_residual (`bool`, *optional*, defaults to `True`): Whether to use bias in the pre-activate residual units of the fusion blocks. add_projection (`bool`, *optional*, defaults to `False`): Whether to add a projection layer before the depth estimation head. use_auxiliary_head (`bool`, *optional*, defaults to `True`): Whether to use an auxiliary head during training. auxiliary_loss_weight (`float`, *optional*, defaults to 0.4): Weight of the cross-entropy loss of the auxiliary head. semantic_classifier_dropout (`float`, *optional*, defaults to 0.1): The dropout ratio for the semantic classification head. backbone_featmap_shape (`list[int]`, *optional*, defaults to `[1, 1024, 24, 24]`): Used only for the `hybrid` embedding type. The shape of the feature maps of the backbone. neck_ignore_stages (`list[int]`, *optional*, defaults to `[0, 1]`): Used only for the `hybrid` embedding type. The stages of the readout layers to ignore. pooler_output_size (`int`, *optional*): Dimensionality of the pooler layer. If None, defaults to `hidden_size`. pooler_act (`str`, *optional*, defaults to `"tanh"`): The activation function to be used by the pooler. Example: ```python >>> from transformers import DPTModel, DPTConfig >>> # Initializing a DPT dpt-large style configuration >>> configuration = DPTConfig() >>> # Initializing a model from the dpt-large style configuration >>> model = DPTModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ```dptbackbone_configi gelugg{Gz?g-q=FTprojectNg?g?tanhc <VPR9d \R4hVP'd\VP\ 4'dVPP RR4\RRVPRRRRR R R R .ROR .RORR//VBwVnpVPR8wd \R4hMGVPR4fVPe'\RRVP/VBwVnpRVn VP'd VPMRVn VP'd VPM.Vn VP'd VPM VPVn \SV`<!R/VBR#)ignorerz8Readout_type must be one of ['ignore', 'add', 'project'] model_typebitr default_config_typedefault_config_kwargsglobal_paddingsame layer_type bottleneckdepths out_featuresembedding_dynamic_paddingTz>>$..55$$// eD+P , $ 4 4 ,$) ,%f ,i"$B/ ' , , (D &  I- !_``. ZZ # /43G3G3S+P, $ 4 4,, (D &)-D %EI^^^d&A&AY]#=A^^^$"9"9QS=A=T=T=T$"9"9Z^ZjZj ''cx<V^8dQh/S[;R&RS[,;R&S[R,;R&S[R,;R&S[;R&S[S[,R,;R&S[S[,R,;R&S[;R &S[R,;R &S[S[S[,,S[S[S[3,,R,;R &S[S[S[,,S[S[S[3,,R,;R &S[R,;R &S[;R&S[R,;R&S[S[,S[S[R3,,R,;R&S[;R&S[S[S[,,S[S[S[,R3,,;R&S[S[,S[S[R3,,;R&S[;R&S[;R&S[R,;R&S[R,;R&S[;R&S[R,;R&S[;R&S[;R&S[S[,;R&S[S[,S[S[R3,,R,;R&S[S[,S[S[R3,,;R&S[S[,R,;R &S[R,;R!&S[;R"&#)#r3Nnum_hidden_layersnum_attention_headsintermediate_size hidden_acthidden_dropout_probattention_probs_dropout_probinitializer_rangelayer_norm_eps image_size patch_size num_channelsr*qkv_bias.r/r(reassemble_factorsneck_hidden_sizesfusion_hidden_size head_in_index!use_batch_norm_in_fusion_residualuse_bias_in_fusion_residualadd_projectionuse_auxiliary_headauxiliary_loss_weightsemantic_loss_ignore_indexsemantic_classifier_dropoutr0r1r r2 pooler_act)intstrfloatlisttupleboolr,r)format __classdict__s"r9 __annotate__DPTConfig.__annotate__sDEFcz&GHt(IJTz(KLMNt+1OP#(#+"4:QR#STDL(UVd3i%S/1D8>WXd3i%S/1D8=YZ* [\]^Tk _`s)eCHo5<Lab!cdS5[)E#+s2B,CCTefCy5c?2Hgh!ijkl(,d{:mn"&3op qrt *st!&uv!$)wx"'2yz!Ic3h7$>R{|S E#s(O3<}~,,t3:@d )ABCr;)r r0r/r1r2)r= )r#r=g?)`rr )rcirg)rc)-__name__ __module__ __qualname____firstlineno____doc__rr sub_configsr3r>r?r@rArBrCrDrErFrGrHr*rIr/r(rJrKrLrMrNrOrPrQrRrSrTr0r1r r2rUr5__annotate_func____static_attributes____classdictcell__ __classcell__)r8r]s@@r9r r s7 rJ$j1K K$&&($(J.17: ##(N;>J;=J LI H?L!LFT5H!M5:%/3 N&*#&&)/2AR6<6:O%)J (Gr;r N) rmhuggingface_hub.dataclassesrbackbone_utilsrconfiguration_utilsrutilsrauto.configuration_autorr __all__r!r;r9rysP.C3#0,-A( A(.A(H -r;