gZpdZddlmZddlmZddlmZmZeje Z GddeeZ dS)zFocalNet model configuration)PretrainedConfig)logging)BackboneConfigMixin*get_aligned_output_features_output_indicescfeZdZdZdZdddddgdgd gd gd d d ddddddddddddffd ZxZS)FocalNetConfiga This is the configuration class to store the configuration of a [`FocalNetModel`]. It is used to instantiate a FocalNet 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 FocalNet [microsoft/focalnet-tiny](https://huggingface.co/microsoft/focalnet-tiny) architecture. Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the documentation from [`PretrainedConfig`] for more information. Args: image_size (`int`, *optional*, defaults to 224): The size (resolution) of each image. patch_size (`int`, *optional*, defaults to 4): The size (resolution) of each patch in the embeddings layer. num_channels (`int`, *optional*, defaults to 3): The number of input channels. embed_dim (`int`, *optional*, defaults to 96): Dimensionality of patch embedding. use_conv_embed (`bool`, *optional*, defaults to `False`): Whether to use convolutional embedding. The authors noted that using convolutional embedding usually improve the performance, but it's not used by default. hidden_sizes (`List[int]`, *optional*, defaults to `[192, 384, 768, 768]`): Dimensionality (hidden size) at each stage. depths (`list(int)`, *optional*, defaults to `[2, 2, 6, 2]`): Depth (number of layers) of each stage in the encoder. focal_levels (`list(int)`, *optional*, defaults to `[2, 2, 2, 2]`): Number of focal levels in each layer of the respective stages in the encoder. focal_windows (`list(int)`, *optional*, defaults to `[3, 3, 3, 3]`): Focal window size in each layer of the respective stages in the encoder. hidden_act (`str` or `function`, *optional*, defaults to `"gelu"`): The non-linear activation function (function or string) in the encoder. If string, `"gelu"`, `"relu"`, `"selu"` and `"gelu_new"` are supported. mlp_ratio (`float`, *optional*, defaults to 4.0): Ratio of MLP hidden dimensionality to embedding dimensionality. hidden_dropout_prob (`float`, *optional*, defaults to 0.0): The dropout probability for all fully connected layers in the embeddings and encoder. drop_path_rate (`float`, *optional*, defaults to 0.1): Stochastic depth rate. use_layerscale (`bool`, *optional*, defaults to `False`): Whether to use layer scale in the encoder. layerscale_value (`float`, *optional*, defaults to 0.0001): The initial value of the layer scale. use_post_layernorm (`bool`, *optional*, defaults to `False`): Whether to use post layer normalization in the encoder. use_post_layernorm_in_modulation (`bool`, *optional*, defaults to `False`): Whether to use post layer normalization in the modulation layer. normalize_modulator (`bool`, *optional*, defaults to `False`): Whether to normalize the modulator. initializer_range (`float`, *optional*, defaults to 0.02): The standard deviation of the truncated_normal_initializer for initializing all weight matrices. layer_norm_eps (`float`, *optional*, defaults to 1e-05): The epsilon used by the layer normalization layers. encoder_stride (`int`, *optional*, defaults to 32): Factor to increase the spatial resolution by in the decoder head for masked image modeling. out_features (`List[str]`, *optional*): If used as backbone, list of features to output. Can be any of `"stem"`, `"stage1"`, `"stage2"`, etc. (depending on how many stages the model has). If unset and `out_indices` is set, will default to the corresponding stages. If unset and `out_indices` is unset, will default to the last stage. Must be in the same order as defined in the `stage_names` attribute. out_indices (`List[int]`, *optional*): If used as backbone, list of indices of features to output. Can be any of 0, 1, 2, etc. (depending on how many stages the model has). If unset and `out_features` is set, will default to the corresponding stages. If unset and `out_features` is unset, will default to the last stage. Must be in the same order as defined in the `stage_names` attribute. Example: ```python >>> from transformers import FocalNetConfig, FocalNetModel >>> # Initializing a FocalNet microsoft/focalnet-tiny style configuration >>> configuration = FocalNetConfig() >>> # Initializing a model (with random weights) from the microsoft/focalnet-tiny style configuration >>> model = FocalNetModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ```focalnetr`F)ir)rr)rrrr)rrrrgelug@gg?g-C6?g{Gz?gh㈵> Nc tjdi|||_||_||_||_||_||_||_||_ | |_ | |_ | |_ | |_ | |_||_||_||_||_||_||_||_||_dgdt/dt1|jdzDz|_t5|||j\|_|_dS)Nstemcg|]}d|S)stage).0idxs o/var/www/html/ai-engine/env/lib/python3.11/site-packages/transformers/models/focalnet/configuration_focalnet.py z+FocalNetConfig.__init__..s&_&_&_}s}}&_&_&_) out_features out_indices stage_namesr)super__init__ image_size patch_size num_channels embed_dimuse_conv_embed hidden_sizesdepths focal_levels focal_windows hidden_act mlp_ratiohidden_dropout_probdrop_path_rateuse_layerscalelayerscale_valueuse_post_layernorm use_post_layernorm_in_modulationnormalize_modulatorinitializer_rangelayer_norm_epsencoder_striderangelenr r _out_features _out_indices)selfr#r$r%r&r'r(r)r*r+r,r-r.r/r0r1r2r3r4r5r6r7rrkwargs __class__s rr"zFocalNetConfig.__init__ls+6 ""6"""$$(",( (*$"#6 ,, 0"40P-#6 !2,,"8&_&_aT[IYIY\]I]@^@^&_&_&__0Z%;DL\1 1 1 -D---r)__name__ __module__ __qualname____doc__ model_typer" __classcell__)r>s@rrrsN N `J)))||!\\"ll ).!15 5 5 5 5 5 5 5 5 5 rrN) rBconfiguration_utilsrutilsrutils.backbone_utilsrr get_loggerr?loggerrrrrrJs#"333333cccccccc  H % %H H H H H (*:H H H H H r