# 🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨 # This file was automatically generated from src/transformers/models/pp_doclayout_v2/modular_pp_doclayout_v2.py. # Do NOT edit this file manually as any edits will be overwritten by the generation of # the file from the modular. If any change should be done, please apply the change to the # modular_pp_doclayout_v2.py file directly. One of our CI enforces this. # 🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨 # Copyright 2026 The PaddlePaddle Team and The HuggingFace Inc. team. All rights reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. from ...backbone_utils import consolidate_backbone_kwargs_to_config from ...configuration_utils import PreTrainedConfig from ..auto import AutoConfig class PPDocLayoutV2ReadingOrderConfig(PreTrainedConfig): r""" This is the configuration class to store the configuration of a [`PPDocLayoutV2ReadingOrder`]. It is used to instantiate the reading order sub-module of the PP-DocLayoutV2 model. This configuration defines the architecture and hyperparameters specific to the reading order detection task within the larger PP-DocLayoutV2 framework. 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 512): Dimension of the encoder layers and the pooled layer. num_attention_heads (`int`, *optional*, defaults to 8): Number of attention heads for each attention layer. attention_probs_dropout_prob (`float`, *optional*, defaults to 0.1): The dropout ratio for the attention probabilities. has_relative_attention_bias (`bool`, *optional*, defaults to `True`): Whether or not to use a relative attention bias in the self-attention mechanism. has_spatial_attention_bias (`bool`, *optional*, defaults to `True`): Whether or not to use a spatial attention bias in the self-attention mechanism. layer_norm_eps (`float`, *optional*, defaults to 1e-05): The epsilon used by the layer normalization layers. hidden_dropout_prob (`float`, *optional*, defaults to 0.1): The dropout probability for all fully connected layers in the embeddings, encoder, and pooler. intermediate_size (`int`, *optional*, defaults to 2048): Dimension of the "intermediate" (i.e., feed-forward) layer in the Transformer encoder. 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"`, `"silu"` and `"gelu_new"` are supported. num_hidden_layers (`int`, *optional*, defaults to 6): Number of the hidden layers. rel_pos_bins (`int`, *optional*, defaults to 32): The number of relative position bins to be used in the self-attention mechanism. max_rel_pos (`int`, *optional*, defaults to 128): The maximum number of relative positions to be used in the self-attention mechanism. rel_2d_pos_bins (`int`, *optional*, defaults to 64): The number of 2D relative position bins in the self-attention mechanism. max_rel_2d_pos (`int`, *optional*, defaults to 256): The maximum number of relative 2D positions in the self-attention mechanism. max_position_embeddings (`int`, *optional*, defaults to 514): 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). max_2d_position_embeddings (`int`, *optional*, defaults to 1024): The maximum value that the 2D position embedding might ever be used with. Typically set this to something large just in case (e.g., 1024). type_vocab_size (`int`, *optional*, defaults to 1): The vocabulary size of the `token_type_ids`. vocab_size (`int`, *optional*, defaults to 4): Vocabulary size of the model. Defines the number of different tokens that can be represented by the `inputs_ids`. initializer_range (`float`, *optional*, defaults to 0.01): The standard deviation of the truncated_normal_initializer for initializing all weight matrices. start_token_id (`int`, *optional*, defaults to 0): Token id representing the start of a sequence. pad_token_id (`int`, *optional*, defaults to 1): Token id used for padding the input sequences. end_token_id (`int`, *optional*, defaults to 2): Token id representing the end of a sequence. pred_token_id (`int`, *optional*, defaults to 3): Token id representing valid prediction positions (placeholders) in the sequence. coordinate_size (`int`, *optional*, defaults to 171): Dimension of the coordinate embeddings. shape_size (`int`, *optional*, defaults to 170): Dimension of the width and height embeddings. num_classes (`int`, *optional*, defaults to 20): Number of labels or classes for the layout elements. relation_bias_embed_dim (`int`, *optional*, defaults to 16): Embedding dimension for the relation bias. relation_bias_theta (`float`, *optional*, defaults to 10000): Temperature parameter used for relation bias scaling. relation_bias_scale (`float`, *optional*, defaults to 100): Scale parameter for the relation bias. global_pointer_head_size (`int`, *optional*, defaults to 64): The size of the global pointer head. gp_dropout_value (`float`, *optional*, defaults to 0.0): The dropout probability in the global pointer head. """ def __init__( self, hidden_size=512, num_attention_heads=8, attention_probs_dropout_prob=0.1, has_relative_attention_bias=False, has_spatial_attention_bias=True, layer_norm_eps=1e-5, hidden_dropout_prob=0.1, intermediate_size=2048, hidden_act="gelu", num_hidden_layers=6, rel_pos_bins=32, max_rel_pos=128, rel_2d_pos_bins=64, max_rel_2d_pos=256, max_position_embeddings=514, max_2d_position_embeddings=1024, type_vocab_size=1, vocab_size=4, initializer_range=0.01, start_token_id=0, pad_token_id=1, end_token_id=2, pred_token_id=3, coordinate_size=171, shape_size=170, num_classes=20, relation_bias_embed_dim=16, relation_bias_theta=10000, relation_bias_scale=100, global_pointer_head_size=64, gp_dropout_value=0.0, **kwargs, ): self.hidden_size = hidden_size self.num_attention_heads = num_attention_heads self.attention_probs_dropout_prob = attention_probs_dropout_prob self.has_relative_attention_bias = has_relative_attention_bias self.has_spatial_attention_bias = has_spatial_attention_bias self.layer_norm_eps = layer_norm_eps self.hidden_dropout_prob = hidden_dropout_prob self.intermediate_size = intermediate_size self.hidden_act = hidden_act self.num_hidden_layers = num_hidden_layers self.rel_pos_bins = rel_pos_bins self.max_rel_pos = max_rel_pos self.rel_2d_pos_bins = rel_2d_pos_bins self.max_rel_2d_pos = max_rel_2d_pos self.max_position_embeddings = max_position_embeddings self.max_2d_position_embeddings = max_2d_position_embeddings self.type_vocab_size = type_vocab_size self.vocab_size = vocab_size self.initializer_range = initializer_range self.start_token_id = start_token_id self.pad_token_id = pad_token_id self.end_token_id = end_token_id self.pred_token_id = pred_token_id self.coordinate_size = coordinate_size self.shape_size = shape_size self.num_classes = num_classes self.relation_bias_embed_dim = relation_bias_embed_dim self.relation_bias_theta = relation_bias_theta self.relation_bias_scale = relation_bias_scale self.global_pointer_head_size = global_pointer_head_size self.gp_dropout_value = gp_dropout_value super().__init__(**kwargs) class PPDocLayoutV2Config(PreTrainedConfig): r""" This is the configuration class to store the configuration of a [`PP-DocLayoutV2`]. It is used to instantiate a PP-DocLayoutV2 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 PP-DocLayoutV2 [PaddlePaddle/PP-DocLayoutV2_safetensors](https://huggingface.co/PaddlePaddle/PP-DocLayoutV2_safetensors) architecture. Configuration objects inherit from [`PreTrainedConfig`] and can be used to control the model outputs. Read the documentation from [`PreTrainedConfig`] for more information. Args: initializer_range (`float`, *optional*, defaults to 0.01): The standard deviation of the truncated_normal_initializer for initializing all weight matrices. initializer_bias_prior_prob (`float`, *optional*): The prior probability used by the bias initializer to initialize biases for `enc_score_head` and `class_embed`. If `None`, `prior_prob` computed as `prior_prob = 1 / (num_labels + 1)` while initializing model weights. layer_norm_eps (`float`, *optional*, defaults to 1e-05): The epsilon used by the layer normalization layers. batch_norm_eps (`float`, *optional*, defaults to 1e-05): The epsilon used by the batch normalization layers. backbone_config (`Union[dict, "PreTrainedConfig"]`, *optional*, defaults to `RTDetrResNetConfig()`): The configuration of the backbone model. freeze_backbone_batch_norms (`bool`, *optional*, defaults to `True`): Whether to freeze the batch normalization layers in the backbone. encoder_hidden_dim (`int`, *optional*, defaults to 256): Dimension of the layers in hybrid encoder. encoder_in_channels (`list`, *optional*, defaults to `[512, 1024, 2048]`): Multi level features input for encoder. feat_strides (`list[int]`, *optional*, defaults to `[8, 16, 32]`): Strides used in each feature map. encoder_layers (`int`, *optional*, defaults to 1): Total of layers to be used by the encoder. encoder_ffn_dim (`int`, *optional*, defaults to 1024): Dimension of the "intermediate" (often named feed-forward) layer in decoder. encoder_attention_heads (`int`, *optional*, defaults to 8): Number of attention heads for each attention layer in the Transformer encoder. dropout (`float`, *optional*, defaults to 0.0): The ratio for all dropout layers. activation_dropout (`float`, *optional*, defaults to 0.0): The dropout ratio for activations inside the fully connected layer. encode_proj_layers (`list[int]`, *optional*, defaults to `[2]`): Indexes of the projected layers to be used in the encoder. positional_encoding_temperature (`int`, *optional*, defaults to 10000): The temperature parameter used to create the positional encodings. encoder_activation_function (`str`, *optional*, defaults to `"gelu"`): The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`, `"relu"`, `"silu"` and `"gelu_new"` are supported. activation_function (`str`, *optional*, defaults to `"silu"`): The non-linear activation function (function or string) in the general layer. If string, `"gelu"`, `"relu"`, `"silu"` and `"gelu_new"` are supported. eval_size (`tuple[int, int]`, *optional*): Height and width used to computes the effective height and width of the position embeddings after taking into account the stride. normalize_before (`bool`, *optional*, defaults to `False`): Determine whether to apply layer normalization in the transformer encoder layer before self-attention and feed-forward modules. hidden_expansion (`float`, *optional*, defaults to 1.0): Expansion ratio to enlarge the dimension size of RepVGGBlock and CSPRepLayer. d_model (`int`, *optional*, defaults to 256): Dimension of the layers exclude hybrid encoder. num_queries (`int`, *optional*, defaults to 300): Number of object queries. decoder_in_channels (`list`, *optional*, defaults to `[256, 256, 256]`): Multi level features dimension for decoder decoder_ffn_dim (`int`, *optional*, defaults to 1024): Dimension of the "intermediate" (often named feed-forward) layer in decoder. num_feature_levels (`int`, *optional*, defaults to 3): The number of input feature levels. decoder_n_points (`int`, *optional*, defaults to 4): The number of sampled keys in each feature level for each attention head in the decoder. decoder_layers (`int`, *optional*, defaults to 6): Number of decoder layers. decoder_attention_heads (`int`, *optional*, defaults to 8): Number of attention heads for each attention layer in the Transformer decoder. decoder_activation_function (`str`, *optional*, defaults to `"relu"`): The non-linear activation function (function or string) in the decoder. If string, `"gelu"`, `"relu"`, `"silu"` and `"gelu_new"` are supported. attention_dropout (`float`, *optional*, defaults to 0.0): The dropout ratio for the attention probabilities. num_denoising (`int`, *optional*, defaults to 100): The total number of denoising tasks or queries to be used for contrastive denoising. label_noise_ratio (`float`, *optional*, defaults to 0.5): The fraction of denoising labels to which random noise should be added. box_noise_scale (`float`, *optional*, defaults to 1.0): Scale or magnitude of noise to be added to the bounding boxes. learn_initial_query (`bool`, *optional*, defaults to `False`): Indicates whether the initial query embeddings for the decoder should be learned during training anchor_image_size (`tuple[int, int]`, *optional*): Height and width of the input image used during evaluation to generate the bounding box anchors. If None, automatic generate anchor is applied. disable_custom_kernels (`bool`, *optional*, defaults to `True`): Whether to disable custom kernels. is_encoder_decoder (`bool`, *optional*, defaults to `True`): Whether the architecture has an encoder decoder structure. class_thresholds (`list[float]`, *optional*): The thresholds for each label. class_order (`list[int]`, *optional*): The priority for each label. reading_order_config (`dict`, *optional*): The configuration of a `PPDocLayoutV2ReadingOrder`. Examples: ```python >>> from transformers import PPDocLayoutV2Config, PPDocLayoutV2ForObjectDetection >>> # Initializing a PP-DocLayoutV2 configuration >>> configuration = PPDocLayoutV2Config() >>> # Initializing a model (with random weights) from the configuration >>> model = PPDocLayoutV2ForObjectDetection(configuration) >>> # Accessing the model configuration >>> configuration = model.config ```""" model_type = "pp_doclayout_v2" sub_configs = {"backbone_config": AutoConfig, "reading_order_config": PPDocLayoutV2ReadingOrderConfig} layer_types = ("basic", "bottleneck") attribute_map = { "hidden_size": "d_model", "num_attention_heads": "encoder_attention_heads", } def __init__( self, initializer_range=0.01, initializer_bias_prior_prob=None, layer_norm_eps=1e-5, batch_norm_eps=1e-5, # backbone backbone_config=None, freeze_backbone_batch_norms=True, # encoder HybridEncoder encoder_hidden_dim=256, encoder_in_channels=[512, 1024, 2048], feat_strides=[8, 16, 32], encoder_layers=1, encoder_ffn_dim=1024, encoder_attention_heads=8, dropout=0.0, activation_dropout=0.0, encode_proj_layers=[2], positional_encoding_temperature=10000, encoder_activation_function="gelu", activation_function="silu", eval_size=None, normalize_before=False, hidden_expansion=1.0, # decoder PPDocLayoutV2Transformer d_model=256, num_queries=300, decoder_in_channels=[256, 256, 256], decoder_ffn_dim=1024, num_feature_levels=3, decoder_n_points=4, decoder_layers=6, decoder_attention_heads=8, decoder_activation_function="relu", attention_dropout=0.0, num_denoising=100, label_noise_ratio=0.5, box_noise_scale=1.0, learn_initial_query=False, anchor_image_size=None, disable_custom_kernels=True, is_encoder_decoder=True, # label class_thresholds=None, class_order=None, reading_order_config=None, **kwargs, ): self.initializer_range = initializer_range self.initializer_bias_prior_prob = initializer_bias_prior_prob self.layer_norm_eps = layer_norm_eps self.batch_norm_eps = batch_norm_eps if isinstance(reading_order_config, dict): self.reading_order_config = self.sub_configs["reading_order_config"](**reading_order_config) elif reading_order_config is None: self.reading_order_config = self.sub_configs["reading_order_config"]() backbone_config, kwargs = consolidate_backbone_kwargs_to_config( backbone_config=backbone_config, default_config_type="hgnet_v2", default_config_kwargs={ "arch": "L", "return_idx": [1, 2, 3], "freeze_stem_only": True, "freeze_at": 0, "freeze_norm": True, "lr_mult_list": [0, 0.05, 0.05, 0.05, 0.05], "out_features": ["stage2", "stage3", "stage4"], }, **kwargs, ) self.backbone_config = backbone_config self.freeze_backbone_batch_norms = freeze_backbone_batch_norms # ---- encoder ---- self.encoder_hidden_dim = encoder_hidden_dim self.encoder_in_channels = list(encoder_in_channels) self.feat_strides = list(feat_strides) self.encoder_layers = encoder_layers self.encoder_ffn_dim = encoder_ffn_dim self.encoder_attention_heads = encoder_attention_heads self.dropout = dropout self.activation_dropout = activation_dropout self.encode_proj_layers = list(encode_proj_layers) self.positional_encoding_temperature = positional_encoding_temperature self.encoder_activation_function = encoder_activation_function self.activation_function = activation_function self.eval_size = list(eval_size) if eval_size is not None else None self.normalize_before = normalize_before self.hidden_expansion = hidden_expansion # ---- decoder ---- self.d_model = d_model self.num_queries = num_queries self.decoder_in_channels = list(decoder_in_channels) self.decoder_ffn_dim = decoder_ffn_dim self.num_feature_levels = num_feature_levels self.decoder_n_points = decoder_n_points self.decoder_layers = decoder_layers self.decoder_attention_heads = decoder_attention_heads self.decoder_activation_function = decoder_activation_function self.attention_dropout = attention_dropout self.num_denoising = num_denoising self.label_noise_ratio = label_noise_ratio self.box_noise_scale = box_noise_scale self.learn_initial_query = learn_initial_query self.anchor_image_size = list(anchor_image_size) if anchor_image_size is not None else None self.disable_custom_kernels = disable_custom_kernels self.class_thresholds = class_thresholds self.class_order = class_order super().__init__(is_encoder_decoder=is_encoder_decoder, **kwargs) __all__ = ["PPDocLayoutV2Config"]