^jdZddlmZddlmZddlmZedeGdd eZedeGd d eZedeGd d eZ edeGddeZ gdZ y)zSAM model configuration)strict)PreTrainedConfig)auto_docstringzfacebook/sam-vit-huge) checkpointceZdZUdZdZdZeed<dZee eze eefzed<dZ ee eze eefzed<dZ eed <d Z eed <d Zeed <dZeed<fdZxZS)SamPromptEncoderConfigz mask_input_channels (`int`, *optional*, defaults to 16): The number of channels to be fed to the `MaskDecoder` module. num_point_embeddings (`int`, *optional*, defaults to 4): The number of point embeddings to be used. prompt_encoder_config hidden_size image_size patch_sizemask_input_channelsnum_point_embeddingsgelu hidden_actư>layer_norm_epsc `|j|jz|_t|di|yN)rrimage_embedding_sizesuper __post_init__selfkwargs __class__s t/var/www/ramen.bs-engineer-server.com/venv/lib/python3.12/site-packages/transformers/models/sam/configuration_sam.pyrz$SamPromptEncoderConfig.__post_init__*s($(OOt$F! '')__name__ __module__ __qualname____doc__base_config_keyr int__annotations__rlisttuplerrrrstrrfloatr __classcell__r!s@r"r r s.OK48Jd3i%S/1846Jd3i%S/16!! !#!J NE ((r#r ceZdZUdZdZdZeed<dZe ed<dZ eed<d Z eed <d Z eed <d Z eed <dZeed<dZeed<dZeed<dZeed<y)SamMaskDecoderConfiga mlp_dim (`int`, *optional*, defaults to 2048): Dimensionality of the "intermediate" (i.e., feed-forward) layer in the Transformer encoder. attention_downsample_rate (`int`, *optional*, defaults to 2): The downsampling rate of the attention layer. num_multimask_outputs (`int`, *optional*, defaults to 3): The number of outputs from the `SamMaskDecoder` module. In the Segment Anything paper, this is set to 3. iou_head_depth (`int`, *optional*, defaults to 3): The number of layers in the IoU head module. iou_head_hidden_dim (`int`, *optional*, defaults to 256): The dimensionality of the hidden states in the IoU head module. mask_decoder_configr r relurimlp_dimnum_hidden_layersnum_attention_headsattention_downsample_raternum_multimask_outputsiou_head_depthiou_head_hidden_dimrrN)r$r%r&r'r(r r)r*rr-r5r7r9r:r;r<r=rr.rr#r"r2r2/sw ,OKJGSs  %&s&!"3"NC"" NE r#r2ceZdZUdZdZdZdZeed<dZ eed<dZ eed <dZ eed <d Z eed <d Z eeezeeefzed<dZeeezeeefzed<dZeed<dZeed<dZeezed<dZeed<dZeed<dZeed<dZeed<dZeed<dZeed <d!Zeeeed"fzed#<d$Zeed%<d&Zed&zed'<fd(Z xZ!S))SamVisionConfiga output_channels (`int`, *optional*, defaults to 256): Dimensionality of the output channels in the Patch Encoder. use_rel_pos (`bool`, *optional*, defaults to `True`): Whether to use relative position embedding. window_size (`int`, *optional*, defaults to 14): Window size for relative position. global_attn_indexes (`list[int]`, *optional*, defaults to `[2, 5, 8, 11]`): The indexes of the global attention layers. num_pos_feats (`int`, *optional*, defaults to 128): The dimensionality of the position embedding. mlp_dim (`int`, *optional*): The dimensionality of the MLP layer in the Transformer encoder. If `None`, defaults to `mlp_ratio * hidden_size`. Example: ```python >>> from transformers import ( ... SamVisionConfig, ... SamVisionModel, ... ) >>> # Initializing a SamVisionConfig with `"facebook/sam-vit-huge"` style configuration >>> configuration = SamVisionConfig() >>> # Initializing a SamVisionModel (with random weights) from the `"facebook/sam-vit-huge"` style configuration >>> model = SamVisionModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ``` vision_configsam_vision_modelir r output_channels r7r9r num_channelsr rrrrrrrgattention_dropoutg|=initializer_rangeTqkv_biasg@ mlp_ratio use_abs_pos use_rel_pos window_size)r6r8 .global_attn_indexes num_pos_featsNr5c |j"t|j|jzn |j|_|jdz|_t |di|y)Nr6r)r5r)r rHscalerrrs r"rzSamVisionConfig.__post_init__sRAEAUs4++dnn<=[_[g[g %%*  ''r#)"r$r%r&r'r( model_typer r)r*rBr7r9rDrr+r,rrr-rr.rErFrGboolrHrIrJrLrOrQr5rr/r0s@r"r?r?Ms% B&O#JKOSs!!L#48Jd3i%S/1846Jd3i%S/16J!NE!%(us{($u$HdIuKKK7DcU38_4DM3GS4Z((r#r?ceZdZUdZdZeeedZdZ e e zdze d<dZ e e zdze d<dZe e zdze d<dZee d <d Zee d <fd ZxZS) SamConfiga; prompt_encoder_config (Union[`dict`, `SamPromptEncoderConfig`], *optional*): Dictionary of configuration options used to initialize [`SamPromptEncoderConfig`]. mask_decoder_config (Union[`dict`, `SamMaskDecoderConfig`], *optional*): Dictionary of configuration options used to initialize [`SamMaskDecoderConfig`]. Example: ```python >>> from transformers import ( ... SamVisionConfig, ... SamPromptEncoderConfig, ... SamMaskDecoderConfig, ... SamModel, ... ) >>> # Initializing a SamConfig with `"facebook/sam-vit-huge"` style configuration >>> configuration = SamConfig() >>> # Initializing a SamModel (with random weights) from the `"facebook/sam-vit-huge"` style configuration >>> model = SamModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config >>> # We can also initialize a SamConfig from a SamVisionConfig, SamPromptEncoderConfig, and SamMaskDecoderConfig >>> # Initializing SAM vision, SAM Q-Former and language model configurations >>> vision_config = SamVisionConfig() >>> prompt_encoder_config = SamPromptEncoderConfig() >>> mask_decoder_config = SamMaskDecoderConfig() >>> config = SamConfig(vision_config, prompt_encoder_config, mask_decoder_config) ```sam)r r3r@Nr@r r3g{Gz?rFTtie_word_embeddingsc t|jtrtdi|j|_n|jt|_t|jtrt di|j|_n|jt |_t|j trtdi|j |_n|j t|_t|$di|yr) isinstancer@dictr?r r r3r2rrrs r"rzSamConfig.__post_init__s d(($ /!0!F43E3E!FD     '!0!2D  d00$ 7)?)]$B\B\)]D &  ' ' /)?)AD & d.. 5';'Wd>V>V'WD $  % % -';'=D $ ''r#)r$r%r&r'rTr r2r? sub_configsr@r\rr*r r3rFr.rYrUrr/r0s@r"rWrWs! FJ!73(K 59M4**T18<@4"22T9@:> 0047>#u# $$((r#rW)rWr2r r?N) r'huggingface_hub.dataclassesrconfiguration_utilsrutilsrr r2r?rW__all__rr#r"rbs.3#23(-(4(.23!+!4!823<(&<(4<(~23A( A(4A(H ]r#