BROS
Overview
BROS モデルは、Teakgyu Hon、Donghyun Kim、Mingi Ji, Wonseok Hwang, Daehyun Nam, Sungrae Park によって BROS: A Pre-trained Language Model Focusing on Text and Layout for Better Key Information Extraction from Documents で提案されました。
BROS は BERT Relying On Spatality の略です。これは、一連のトークンとその境界ボックスを入力として受け取り、一連の隠れ状態を出力するエンコーダー専用の Transformer モデルです。 BROS は、絶対的な空間情報を使用する代わりに、相対的な空間情報をエンコードします。
BERT で使用されるトークンマスク言語モデリング目標 (TMLM) と新しいエリアマスク言語モデリング目標 (AMLM) の 2 つの目標で事前トレーニングされています。 TMLM では、トークンはランダムにマスクされ、モデルは空間情報と他のマスクされていないトークンを使用してマスクされたトークンを予測します。 AMLM は TMLM の 2D バージョンです。テキスト トークンをランダムにマスクし、TMLM と同じ情報で予測しますが、テキスト ブロック (領域) をマスクします。
BrosForTokenClassificationには、BrosModel の上に単純な線形層があります。各トークンのラベルを予測します。
BrosSpadeEEForTokenClassificationには、BrosModel の上にinitial_token_classifierとsubsequent_token_classifierがあります。 initial_token_classifier は各エンティティの最初のトークンを予測するために使用され、subsequent_token_classifier はエンティティ内の次のトークンを予測するために使用されます。 BrosSpadeELForTokenClassificationには BrosModel の上にentity_linkerがあります。 entity_linker は 2 つのエンティティ間の関係を予測するために使用されます。
BrosForTokenClassificationとBrosSpadeEEForTokenClassificationは基本的に同じジョブを実行します。ただし、BrosForTokenClassificationは入力トークンが完全にシリアル化されていることを前提としています (トークンは 2D 空間に存在するため、これは非常に困難な作業です)。一方、BrosSpadeEEForTokenClassificationは 1 つのトークンから次の接続トークンを予測するため、シリアル化エラーの処理をより柔軟に行うことができます。
BrosSpadeELForTokenClassification はエンティティ内のリンク タスクを実行します。これら 2 つのエンティティが何らかの関係を共有する場合、(あるエンティティの) 1 つのトークンから (別のエンティティの) 別のトークンへの関係を予測します。
BROS は、明示的な視覚機能に依存せずに、FUNSD、SROIE、CORD、SciTSR などの Key Information Extraction (KIE) ベンチマークで同等以上の結果を達成します。
論文の要約は次のとおりです。
文書画像からの重要情報抽出 (KIE) には、2 次元 (2D) 空間におけるテキストの文脈的および空間的意味論を理解する必要があります。最近の研究の多くは、文書画像の視覚的特徴とテキストおよびそのレイアウトを組み合わせることに重点を置いた事前トレーニング済み言語モデルを開発することで、この課題を解決しようとしています。一方、このペーパーでは、テキストとレイアウトの効果的な組み合わせという基本に立ち返ってこの問題に取り組みます。具体的には、BROS (BERT Relying On Spatality) という名前の事前トレーニング済み言語モデルを提案します。この言語モデルは、2D 空間内のテキストの相対位置をエンコードし、エリア マスキング戦略を使用してラベルのないドキュメントから学習します。 2D 空間内のテキストを理解するためのこの最適化されたトレーニング スキームにより、BROS は、視覚的な特徴に依存することなく、4 つの KIE ベンチマーク (FUNSD、SROIE、CORD、および SciTSR) で以前の方法と比較して同等以上のパフォーマンスを示しました。また、この論文では、KIE タスクにおける 2 つの現実世界の課題 ((1) 間違ったテキスト順序によるエラーの最小化、および (2) 少数の下流例からの効率的な学習) を明らかにし、以前の方法に対する BROS の優位性を実証します。*
このモデルは jinho8345 によって寄稿されました。元のコードは ここ にあります。
Usage tips and examples
- forward() には、
input_idsとbbox(バウンディング ボックス) が必要です。各境界ボックスは、(x0、y0、x1、y1) 形式 (左上隅、右下隅) である必要があります。境界ボックスの取得は外部 OCR システムに依存します。 「x」座標はドキュメント画像の幅で正規化する必要があり、「y」座標はドキュメント画像の高さで正規化する必要があります。
def expand_and_normalize_bbox(bboxes, doc_width, doc_height):
# here, bboxes are numpy array
# Normalize bbox -> 0 ~ 1
bboxes[:, [0, 2]] = bboxes[:, [0, 2]] / width
bboxes[:, [1, 3]] = bboxes[:, [1, 3]] / height- [
~transformers.BrosForTokenClassification.forward、~transformers.BrosSpadeEEForTokenClassification.forward、~transformers.BrosSpadeEEForTokenClassification.forward] では、損失計算にinput_idsとbboxだけでなくbox_first_token_maskも必要です。これは、各ボックスの先頭以外のトークンを除外するためのマスクです。このマスクは、単語からinput_idsを作成するときに境界ボックスの開始トークン インデックスを保存することで取得できます。次のコードでbox_first_token_maskを作成できます。
def make_box_first_token_mask(bboxes, words, tokenizer, max_seq_length=512):
box_first_token_mask = np.zeros(max_seq_length, dtype=np.bool_)
# encode(tokenize) each word from words (List[str])
input_ids_list: List[List[int]] = [tokenizer.encode(e, add_special_tokens=False) for e in words]
# get the length of each box
tokens_length_list: List[int] = [len(l) for l in input_ids_list]
box_end_token_indices = np.array(list(itertools.accumulate(tokens_length_list)))
box_start_token_indices = box_end_token_indices - np.array(tokens_length_list)
# filter out the indices that are out of max_seq_length
box_end_token_indices = box_end_token_indices[box_end_token_indices < max_seq_length - 1]
if len(box_start_token_indices) > len(box_end_token_indices):
box_start_token_indices = box_start_token_indices[: len(box_end_token_indices)]
# set box_start_token_indices to True
box_first_token_mask[box_start_token_indices] = True
return box_first_token_mask
Resources
- デモ スクリプトは こちら にあります。
BrosConfig
class transformers.BrosConfig
< source >( vocab_size = 30522 hidden_size = 768 num_hidden_layers = 12 num_attention_heads = 12 intermediate_size = 3072 hidden_act = 'gelu' hidden_dropout_prob = 0.1 attention_probs_dropout_prob = 0.1 max_position_embeddings = 512 type_vocab_size = 2 initializer_range = 0.02 layer_norm_eps = 1e-12 pad_token_id = 0 dim_bbox = 8 bbox_scale = 100.0 n_relations = 1 classifier_dropout_prob = 0.1 **kwargs )
Parameters
- vocab_size (
int, optional, defaults to 30522) — Vocabulary size of the Bros model. Defines the number of different tokens that can be represented by theinputs_idspassed when calling BrosModel orTFBrosModel. - hidden_size (
int, optional, defaults to 768) — Dimensionality of the encoder layers and the pooler layer. - num_hidden_layers (
int, optional, defaults to 12) — Number of hidden layers in the Transformer encoder. - num_attention_heads (
int, optional, defaults to 12) — Number of attention heads for each attention layer in the Transformer encoder. - intermediate_size (
int, optional, defaults to 3072) — Dimensionality of the “intermediate” (often named feed-forward) layer in the Transformer encoder. - hidden_act (
strorCallable, 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. - hidden_dropout_prob (
float, optional, defaults to 0.1) — The dropout probability for all fully connected layers in the embeddings, encoder, and pooler. - attention_probs_dropout_prob (
float, optional, defaults to 0.1) — The dropout ratio for the attention probabilities. - max_position_embeddings (
int, optional, defaults to 512) — 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). - type_vocab_size (
int, optional, defaults to 2) — The vocabulary size of thetoken_type_idspassed when calling BrosModel orTFBrosModel. - 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-12) — The epsilon used by the layer normalization layers. - pad_token_id (
int, optional, defaults to 0) — The index of the padding token in the token vocabulary. - dim_bbox (
int, optional, defaults to 8) — The dimension of the bounding box coordinates. (x0, y1, x1, y0, x1, y1, x0, y1) - bbox_scale (
float, optional, defaults to 100.0) — The scale factor of the bounding box coordinates. - n_relations (
int, optional, defaults to 1) — The number of relations for SpadeEE(entity extraction), SpadeEL(entity linking) head. - classifier_dropout_prob (
float, optional, defaults to 0.1) — The dropout ratio for the classifier head.
This is the configuration class to store the configuration of a BrosModel or a TFBrosModel. It is used to
instantiate a Bros 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 Bros
jinho8345/bros-base-uncased architecture.
Configuration objects inherit from PretrainedConfig and can be used to control the model outputs. Read the documentation from PretrainedConfig for more information.
Examples:
>>> from transformers import BrosConfig, BrosModel
>>> # Initializing a BROS jinho8345/bros-base-uncased style configuration
>>> configuration = BrosConfig()
>>> # Initializing a model from the jinho8345/bros-base-uncased style configuration
>>> model = BrosModel(configuration)
>>> # Accessing the model configuration
>>> configuration = model.configBrosProcessor
class transformers.BrosProcessor
< source >( tokenizer = None **kwargs )
Constructs a Bros processor which wraps a BERT tokenizer.
BrosProcessor offers all the functionalities of BertTokenizerFast. See the docstring of
call() and decode() for more information.
__call__
< source >( text: Union = None add_special_tokens: bool = True padding: Union = False truncation: Union = None max_length: Optional = None stride: int = 0 pad_to_multiple_of: Optional = None return_token_type_ids: Optional = None return_attention_mask: Optional = None return_overflowing_tokens: bool = False return_special_tokens_mask: bool = False return_offsets_mapping: bool = False return_length: bool = False verbose: bool = True return_tensors: Union = None **kwargs )
This method uses BertTokenizerFast.call() to prepare text for the model.
Please refer to the docstring of the above two methods for more information.
BrosModel
class transformers.BrosModel
< source >( config add_pooling_layer = True )
Parameters
- config (BrosConfig) — Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.
The bare Bros Model transformer outputting raw hidden-states without any specific head on top. This model is also a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.
forward
< source >( input_ids: Optional = None bbox: Optional = None attention_mask: Optional = None token_type_ids: Optional = None position_ids: Optional = None head_mask: Optional = None inputs_embeds: Optional = None encoder_hidden_states: Optional = None encoder_attention_mask: Optional = None past_key_values: Optional = None use_cache: Optional = None output_attentions: Optional = None output_hidden_states: Optional = None return_dict: Optional = None ) → transformers.modeling_outputs.BaseModelOutputWithPoolingAndCrossAttentions or tuple(torch.FloatTensor)
Parameters
- input_ids (
torch.LongTensorof shape(batch_size, sequence_length)) — Indices of input sequence tokens in the vocabulary.Indices can be obtained using BrosProcessor. See PreTrainedTokenizer.encode() and PreTrainedTokenizer.call() for details.
- bbox (‘torch.FloatTensor’ of shape ‘(batch_size, num_boxes, 4)’) — Bounding box coordinates for each token in the input sequence. Each bounding box is a list of four values (x1, y1, x2, y2), where (x1, y1) is the top left corner, and (x2, y2) is the bottom right corner of the bounding box.
- attention_mask (
torch.FloatTensorof shape(batch_size, sequence_length), optional) — Mask to avoid performing attention on padding token indices. Mask values selected in[0, 1]:- 1 for tokens that are not masked,
- 0 for tokens that are masked.
- bbox_first_token_mask (
torch.FloatTensorof shape(batch_size, sequence_length), optional) — Mask to indicate the first token of each bounding box. Mask values selected in[0, 1]:- 1 for tokens that are not masked,
- 0 for tokens that are masked.
- token_type_ids (
torch.LongTensorof shape(batch_size, sequence_length), optional) — Segment token indices to indicate first and second portions of the inputs. Indices are selected in[0, 1]:- 0 corresponds to a sentence A token,
- 1 corresponds to a sentence B token.
- position_ids (
torch.LongTensorof shape(batch_size, sequence_length), optional) — Indices of positions of each input sequence tokens in the position embeddings. Selected in the range[0, config.max_position_embeddings - 1]. - head_mask (
torch.FloatTensorof shape(num_heads,)or(num_layers, num_heads), optional) — Mask to nullify selected heads of the self-attention modules. Mask values selected in[0, 1]:- 1 indicates the head is not masked,
- 0 indicates the head is masked.
- inputs_embeds (
torch.FloatTensorof shape(batch_size, sequence_length, hidden_size), optional) — Optionally, instead of passinginput_idsyou can choose to directly pass an embedded representation. This is useful if you want more control over how to convertinput_idsindices into associated vectors than the model’s internal embedding lookup matrix. - output_attentions (
bool, optional) — Whether or not to return the attentions tensors of all attention layers. Seeattentionsunder returned tensors for more detail. - output_hidden_states (
bool, optional) — Whether or not to return the hidden states of all layers. Seehidden_statesunder returned tensors for more detail. - return_dict (
bool, optional) — Whether or not to return a ModelOutput instead of a plain tuple.
Returns
transformers.modeling_outputs.BaseModelOutputWithPoolingAndCrossAttentions or tuple(torch.FloatTensor)
A transformers.modeling_outputs.BaseModelOutputWithPoolingAndCrossAttentions or a tuple of
torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various
elements depending on the configuration (BrosConfig) and inputs.
-
last_hidden_state (
torch.FloatTensorof shape(batch_size, sequence_length, hidden_size)) — Sequence of hidden-states at the output of the last layer of the model. -
pooler_output (
torch.FloatTensorof shape(batch_size, hidden_size)) — Last layer hidden-state of the first token of the sequence (classification token) after further processing through the layers used for the auxiliary pretraining task. E.g. for BERT-family of models, this returns the classification token after processing through a linear layer and a tanh activation function. The linear layer weights are trained from the next sentence prediction (classification) objective during pretraining. -
hidden_states (
tuple(torch.FloatTensor), optional, returned whenoutput_hidden_states=Trueis passed or whenconfig.output_hidden_states=True) — Tuple oftorch.FloatTensor(one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape(batch_size, sequence_length, hidden_size).Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
-
attentions (
tuple(torch.FloatTensor), optional, returned whenoutput_attentions=Trueis passed or whenconfig.output_attentions=True) — Tuple oftorch.FloatTensor(one for each layer) of shape(batch_size, num_heads, sequence_length, sequence_length).Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.
-
cross_attentions (
tuple(torch.FloatTensor), optional, returned whenoutput_attentions=Trueandconfig.add_cross_attention=Trueis passed or whenconfig.output_attentions=True) — Tuple oftorch.FloatTensor(one for each layer) of shape(batch_size, num_heads, sequence_length, sequence_length).Attentions weights of the decoder’s cross-attention layer, after the attention softmax, used to compute the weighted average in the cross-attention heads.
-
past_key_values (
tuple(tuple(torch.FloatTensor)), optional, returned whenuse_cache=Trueis passed or whenconfig.use_cache=True) — Tuple oftuple(torch.FloatTensor)of lengthconfig.n_layers, with each tuple having 2 tensors of shape(batch_size, num_heads, sequence_length, embed_size_per_head)) and optionally ifconfig.is_encoder_decoder=True2 additional tensors of shape(batch_size, num_heads, encoder_sequence_length, embed_size_per_head).Contains pre-computed hidden-states (key and values in the self-attention blocks and optionally if
config.is_encoder_decoder=Truein the cross-attention blocks) that can be used (seepast_key_valuesinput) to speed up sequential decoding.
The BrosModel forward method, overrides the __call__ special method.
Although the recipe for forward pass needs to be defined within this function, one should call the Module
instance afterwards instead of this since the former takes care of running the pre and post processing steps while
the latter silently ignores them.
Examples:
>>> import torch
>>> from transformers import BrosProcessor, BrosModel
>>> processor = BrosProcessor.from_pretrained("jinho8345/bros-base-uncased")
>>> model = BrosModel.from_pretrained("jinho8345/bros-base-uncased")
>>> encoding = processor("Hello, my dog is cute", add_special_tokens=False, return_tensors="pt")
>>> bbox = torch.tensor([[[0, 0, 1, 1]]]).repeat(1, encoding["input_ids"].shape[-1], 1)
>>> encoding["bbox"] = bbox
>>> outputs = model(**encoding)
>>> last_hidden_states = outputs.last_hidden_stateBrosForTokenClassification
class transformers.BrosForTokenClassification
< source >( config )
Parameters
- config (BrosConfig) — Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.
Bros Model with a token classification head on top (a linear layer on top of the hidden-states output) e.g. for Named-Entity-Recognition (NER) tasks.
This model is also a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.
forward
< source >( input_ids: Optional = None bbox: Optional = None attention_mask: Optional = None bbox_first_token_mask: Optional = None token_type_ids: Optional = None position_ids: Optional = None head_mask: Optional = None inputs_embeds: Optional = None labels: Optional = None output_attentions: Optional = None output_hidden_states: Optional = None return_dict: Optional = None ) → transformers.modeling_outputs.TokenClassifierOutput or tuple(torch.FloatTensor)
Parameters
- input_ids (
torch.LongTensorof shape(batch_size, sequence_length)) — Indices of input sequence tokens in the vocabulary.Indices can be obtained using BrosProcessor. See PreTrainedTokenizer.encode() and PreTrainedTokenizer.call() for details.
- bbox (‘torch.FloatTensor’ of shape ‘(batch_size, num_boxes, 4)’) — Bounding box coordinates for each token in the input sequence. Each bounding box is a list of four values (x1, y1, x2, y2), where (x1, y1) is the top left corner, and (x2, y2) is the bottom right corner of the bounding box.
- attention_mask (
torch.FloatTensorof shape(batch_size, sequence_length), optional) — Mask to avoid performing attention on padding token indices. Mask values selected in[0, 1]:- 1 for tokens that are not masked,
- 0 for tokens that are masked.
- bbox_first_token_mask (
torch.FloatTensorof shape(batch_size, sequence_length), optional) — Mask to indicate the first token of each bounding box. Mask values selected in[0, 1]:- 1 for tokens that are not masked,
- 0 for tokens that are masked.
- token_type_ids (
torch.LongTensorof shape(batch_size, sequence_length), optional) — Segment token indices to indicate first and second portions of the inputs. Indices are selected in[0, 1]:- 0 corresponds to a sentence A token,
- 1 corresponds to a sentence B token.
- position_ids (
torch.LongTensorof shape(batch_size, sequence_length), optional) — Indices of positions of each input sequence tokens in the position embeddings. Selected in the range[0, config.max_position_embeddings - 1]. - head_mask (
torch.FloatTensorof shape(num_heads,)or(num_layers, num_heads), optional) — Mask to nullify selected heads of the self-attention modules. Mask values selected in[0, 1]:- 1 indicates the head is not masked,
- 0 indicates the head is masked.
- inputs_embeds (
torch.FloatTensorof shape(batch_size, sequence_length, hidden_size), optional) — Optionally, instead of passinginput_idsyou can choose to directly pass an embedded representation. This is useful if you want more control over how to convertinput_idsindices into associated vectors than the model’s internal embedding lookup matrix. - output_attentions (
bool, optional) — Whether or not to return the attentions tensors of all attention layers. Seeattentionsunder returned tensors for more detail. - output_hidden_states (
bool, optional) — Whether or not to return the hidden states of all layers. Seehidden_statesunder returned tensors for more detail. - return_dict (
bool, optional) — Whether or not to return a ModelOutput instead of a plain tuple.
Returns
transformers.modeling_outputs.TokenClassifierOutput or tuple(torch.FloatTensor)
A transformers.modeling_outputs.TokenClassifierOutput or a tuple of
torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various
elements depending on the configuration (BrosConfig) and inputs.
-
loss (
torch.FloatTensorof shape(1,), optional, returned whenlabelsis provided) — Classification loss. -
logits (
torch.FloatTensorof shape(batch_size, sequence_length, config.num_labels)) — Classification scores (before SoftMax). -
hidden_states (
tuple(torch.FloatTensor), optional, returned whenoutput_hidden_states=Trueis passed or whenconfig.output_hidden_states=True) — Tuple oftorch.FloatTensor(one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape(batch_size, sequence_length, hidden_size).Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
-
attentions (
tuple(torch.FloatTensor), optional, returned whenoutput_attentions=Trueis passed or whenconfig.output_attentions=True) — Tuple oftorch.FloatTensor(one for each layer) of shape(batch_size, num_heads, sequence_length, sequence_length).Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.
The BrosForTokenClassification forward method, overrides the __call__ special method.
Although the recipe for forward pass needs to be defined within this function, one should call the Module
instance afterwards instead of this since the former takes care of running the pre and post processing steps while
the latter silently ignores them.
Examples:
>>> import torch
>>> from transformers import BrosProcessor, BrosForTokenClassification
>>> processor = BrosProcessor.from_pretrained("jinho8345/bros-base-uncased")
>>> model = BrosForTokenClassification.from_pretrained("jinho8345/bros-base-uncased")
>>> encoding = processor("Hello, my dog is cute", add_special_tokens=False, return_tensors="pt")
>>> bbox = torch.tensor([[[0, 0, 1, 1]]]).repeat(1, encoding["input_ids"].shape[-1], 1)
>>> encoding["bbox"] = bbox
>>> outputs = model(**encoding)BrosSpadeEEForTokenClassification
class transformers.BrosSpadeEEForTokenClassification
< source >( config )
Parameters
- config (BrosConfig) — Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.
Bros Model with a token classification head on top (initial_token_layers and subsequent_token_layer on top of the hidden-states output) e.g. for Named-Entity-Recognition (NER) tasks. The initial_token_classifier is used to predict the first token of each entity, and the subsequent_token_classifier is used to predict the subsequent tokens within an entity. Compared to BrosForTokenClassification, this model is more robust to serialization errors since it predicts next token from one token.
This model is also a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.
forward
< source >( input_ids: Optional = None bbox: Optional = None attention_mask: Optional = None bbox_first_token_mask: Optional = None token_type_ids: Optional = None position_ids: Optional = None head_mask: Optional = None inputs_embeds: Optional = None initial_token_labels: Optional = None subsequent_token_labels: Optional = None output_attentions: Optional = None output_hidden_states: Optional = None return_dict: Optional = None ) → transformers.models.bros.modeling_bros.BrosSpadeOutput or tuple(torch.FloatTensor)
Parameters
- input_ids (
torch.LongTensorof shape(batch_size, sequence_length)) — Indices of input sequence tokens in the vocabulary.Indices can be obtained using BrosProcessor. See PreTrainedTokenizer.encode() and PreTrainedTokenizer.call() for details.
- bbox (‘torch.FloatTensor’ of shape ‘(batch_size, num_boxes, 4)’) — Bounding box coordinates for each token in the input sequence. Each bounding box is a list of four values (x1, y1, x2, y2), where (x1, y1) is the top left corner, and (x2, y2) is the bottom right corner of the bounding box.
- attention_mask (
torch.FloatTensorof shape(batch_size, sequence_length), optional) — Mask to avoid performing attention on padding token indices. Mask values selected in[0, 1]:- 1 for tokens that are not masked,
- 0 for tokens that are masked.
- bbox_first_token_mask (
torch.FloatTensorof shape(batch_size, sequence_length), optional) — Mask to indicate the first token of each bounding box. Mask values selected in[0, 1]:- 1 for tokens that are not masked,
- 0 for tokens that are masked.
- token_type_ids (
torch.LongTensorof shape(batch_size, sequence_length), optional) — Segment token indices to indicate first and second portions of the inputs. Indices are selected in[0, 1]:- 0 corresponds to a sentence A token,
- 1 corresponds to a sentence B token.
- position_ids (
torch.LongTensorof shape(batch_size, sequence_length), optional) — Indices of positions of each input sequence tokens in the position embeddings. Selected in the range[0, config.max_position_embeddings - 1]. - head_mask (
torch.FloatTensorof shape(num_heads,)or(num_layers, num_heads), optional) — Mask to nullify selected heads of the self-attention modules. Mask values selected in[0, 1]:- 1 indicates the head is not masked,
- 0 indicates the head is masked.
- inputs_embeds (
torch.FloatTensorof shape(batch_size, sequence_length, hidden_size), optional) — Optionally, instead of passinginput_idsyou can choose to directly pass an embedded representation. This is useful if you want more control over how to convertinput_idsindices into associated vectors than the model’s internal embedding lookup matrix. - output_attentions (
bool, optional) — Whether or not to return the attentions tensors of all attention layers. Seeattentionsunder returned tensors for more detail. - output_hidden_states (
bool, optional) — Whether or not to return the hidden states of all layers. Seehidden_statesunder returned tensors for more detail. - return_dict (
bool, optional) — Whether or not to return a ModelOutput instead of a plain tuple.
Returns
transformers.models.bros.modeling_bros.BrosSpadeOutput or tuple(torch.FloatTensor)
A transformers.models.bros.modeling_bros.BrosSpadeOutput or a tuple of
torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various
elements depending on the configuration (BrosConfig) and inputs.
-
loss (
torch.FloatTensorof shape(1,), optional, returned whenlabelsis provided) — Classification loss. -
initial_token_logits (
torch.FloatTensorof shape(batch_size, sequence_length, config.num_labels)) — Classification scores for entity initial tokens (before SoftMax). -
subsequent_token_logits (
torch.FloatTensorof shape(batch_size, sequence_length, sequence_length+1)) — Classification scores for entity sequence tokens (before SoftMax). -
hidden_states (
tuple(torch.FloatTensor), optional, returned whenoutput_hidden_states=Trueis passed or whenconfig.output_hidden_states=True) — Tuple oftorch.FloatTensor(one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape(batch_size, sequence_length, hidden_size).Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
-
attentions (
tuple(torch.FloatTensor), optional, returned whenoutput_attentions=Trueis passed or whenconfig.output_attentions=True) — Tuple oftorch.FloatTensor(one for each layer) of shape(batch_size, num_heads, sequence_length, sequence_length).Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.
The BrosSpadeEEForTokenClassification forward method, overrides the __call__ special method.
Although the recipe for forward pass needs to be defined within this function, one should call the Module
instance afterwards instead of this since the former takes care of running the pre and post processing steps while
the latter silently ignores them.
Examples:
>>> import torch
>>> from transformers import BrosProcessor, BrosSpadeEEForTokenClassification
>>> processor = BrosProcessor.from_pretrained("jinho8345/bros-base-uncased")
>>> model = BrosSpadeEEForTokenClassification.from_pretrained("jinho8345/bros-base-uncased")
>>> encoding = processor("Hello, my dog is cute", add_special_tokens=False, return_tensors="pt")
>>> bbox = torch.tensor([[[0, 0, 1, 1]]]).repeat(1, encoding["input_ids"].shape[-1], 1)
>>> encoding["bbox"] = bbox
>>> outputs = model(**encoding)BrosSpadeELForTokenClassification
class transformers.BrosSpadeELForTokenClassification
< source >( config )
Parameters
- config (BrosConfig) — Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.
Bros Model with a token classification head on top (a entity_linker layer on top of the hidden-states output) e.g. for Entity-Linking. The entity_linker is used to predict intra-entity links (one entity to another entity).
This model is also a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.
forward
< source >( input_ids: Optional = None bbox: Optional = None attention_mask: Optional = None bbox_first_token_mask: Optional = None token_type_ids: Optional = None position_ids: Optional = None head_mask: Optional = None inputs_embeds: Optional = None labels: Optional = None output_attentions: Optional = None output_hidden_states: Optional = None return_dict: Optional = None ) → transformers.modeling_outputs.TokenClassifierOutput or tuple(torch.FloatTensor)
Parameters
- input_ids (
torch.LongTensorof shape(batch_size, sequence_length)) — Indices of input sequence tokens in the vocabulary.Indices can be obtained using BrosProcessor. See PreTrainedTokenizer.encode() and PreTrainedTokenizer.call() for details.
- bbox (‘torch.FloatTensor’ of shape ‘(batch_size, num_boxes, 4)’) — Bounding box coordinates for each token in the input sequence. Each bounding box is a list of four values (x1, y1, x2, y2), where (x1, y1) is the top left corner, and (x2, y2) is the bottom right corner of the bounding box.
- attention_mask (
torch.FloatTensorof shape(batch_size, sequence_length), optional) — Mask to avoid performing attention on padding token indices. Mask values selected in[0, 1]:- 1 for tokens that are not masked,
- 0 for tokens that are masked.
- bbox_first_token_mask (
torch.FloatTensorof shape(batch_size, sequence_length), optional) — Mask to indicate the first token of each bounding box. Mask values selected in[0, 1]:- 1 for tokens that are not masked,
- 0 for tokens that are masked.
- token_type_ids (
torch.LongTensorof shape(batch_size, sequence_length), optional) — Segment token indices to indicate first and second portions of the inputs. Indices are selected in[0, 1]:- 0 corresponds to a sentence A token,
- 1 corresponds to a sentence B token.
- position_ids (
torch.LongTensorof shape(batch_size, sequence_length), optional) — Indices of positions of each input sequence tokens in the position embeddings. Selected in the range[0, config.max_position_embeddings - 1]. - head_mask (
torch.FloatTensorof shape(num_heads,)or(num_layers, num_heads), optional) — Mask to nullify selected heads of the self-attention modules. Mask values selected in[0, 1]:- 1 indicates the head is not masked,
- 0 indicates the head is masked.
- inputs_embeds (
torch.FloatTensorof shape(batch_size, sequence_length, hidden_size), optional) — Optionally, instead of passinginput_idsyou can choose to directly pass an embedded representation. This is useful if you want more control over how to convertinput_idsindices into associated vectors than the model’s internal embedding lookup matrix. - output_attentions (
bool, optional) — Whether or not to return the attentions tensors of all attention layers. Seeattentionsunder returned tensors for more detail. - output_hidden_states (
bool, optional) — Whether or not to return the hidden states of all layers. Seehidden_statesunder returned tensors for more detail. - return_dict (
bool, optional) — Whether or not to return a ModelOutput instead of a plain tuple.
Returns
transformers.modeling_outputs.TokenClassifierOutput or tuple(torch.FloatTensor)
A transformers.modeling_outputs.TokenClassifierOutput or a tuple of
torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various
elements depending on the configuration (BrosConfig) and inputs.
-
loss (
torch.FloatTensorof shape(1,), optional, returned whenlabelsis provided) — Classification loss. -
logits (
torch.FloatTensorof shape(batch_size, sequence_length, config.num_labels)) — Classification scores (before SoftMax). -
hidden_states (
tuple(torch.FloatTensor), optional, returned whenoutput_hidden_states=Trueis passed or whenconfig.output_hidden_states=True) — Tuple oftorch.FloatTensor(one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape(batch_size, sequence_length, hidden_size).Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
-
attentions (
tuple(torch.FloatTensor), optional, returned whenoutput_attentions=Trueis passed or whenconfig.output_attentions=True) — Tuple oftorch.FloatTensor(one for each layer) of shape(batch_size, num_heads, sequence_length, sequence_length).Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.
The BrosSpadeELForTokenClassification forward method, overrides the __call__ special method.
Although the recipe for forward pass needs to be defined within this function, one should call the Module
instance afterwards instead of this since the former takes care of running the pre and post processing steps while
the latter silently ignores them.
Examples:
>>> import torch
>>> from transformers import BrosProcessor, BrosSpadeELForTokenClassification
>>> processor = BrosProcessor.from_pretrained("jinho8345/bros-base-uncased")
>>> model = BrosSpadeELForTokenClassification.from_pretrained("jinho8345/bros-base-uncased")
>>> encoding = processor("Hello, my dog is cute", add_special_tokens=False, return_tensors="pt")
>>> bbox = torch.tensor([[[0, 0, 1, 1]]]).repeat(1, encoding["input_ids"].shape[-1], 1)
>>> encoding["bbox"] = bbox
>>> outputs = model(**encoding)