modeling¶
-
class
LukeModel
(vocab_size=50267, 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=514, type_vocab_size=1, entity_vocab_size=500000, entity_emb_size=256, initializer_range=0.02, pad_token_id=1, entity_pad_token_id=0)[源代码]¶ 基类:
paddlenlp.transformers.luke.modeling.LukePretrainedModel
The bare Luke Model transformer outputting raw hidden-states.
This model inherits from
PretrainedModel
. Refer to the superclass documentation for the generic methods.This model is also a Paddle paddle.nn.Layer subclass. Use it as a regular Paddle Layer and refer to the Paddle documentation for all matter related to general usage and behavior.
- 参数
vocab_size (int, optional) -- Vocabulary size of
inputs_ids
inLukeModel
. Also is the vocab size of token embedding matrix. Defines the number of different tokens that can be represented by theinputs_ids
passed when callingLukeModel
. Defaults to 50267.hidden_size (int, optional) -- Dimensionality of the embedding layer, encoder layer and pooler layer. Defaults to
768
.num_hidden_layers (int, optional) -- Number of hidden layers in the Transformer encoder. Defaults to
12
.num_attention_heads (int, optional) -- Number of attention heads for each attention layer in the Transformer encoder. Defaults to
12
.intermediate_size (int, optional) -- Dimensionality of the feed-forward (ff) layer in the encoder. Input tensors to ff layers are firstly projected from
hidden_size
tointermediate_size
, and then projected back tohidden_size
. Typicallyintermediate_size
is larger thanhidden_size
. Defaults to3072
.hidden_act (str, optional) -- The non-linear activation function in the feed-forward layer.
"gelu"
,"relu"
and any other paddle supported activation functions are supported. Defaults to"gelu"
.hidden_dropout_prob (float, optional) -- The dropout probability for all fully connected layers in the embeddings and encoder. Defaults to
0.1
.attention_probs_dropout_prob (float, optional) -- The dropout probability used in MultiHeadAttention in all encoder layers to drop some attention target. Defaults to
0.1
.max_position_embeddings (int, optional) -- The maximum value of the dimensionality of position encoding, which dictates the maximum supported length of an input sequence. Defaults to
514
.type_vocab_size (int, optional) -- The vocabulary size of
token_type_ids
. Defaults to1
.entity_vocab_size (int, optional) -- Vocabulary size of
entity_ids
inLukeModel
. Also is the vocab size of token entity embedding matrix. Defines the number of different entity that can be represented by theentity_ids
passed when callingLukeModel
. Defaults to 500000.entity_emb_size (int, optional) -- Dimensionality of the entity embedding layer Defaults to
256
.initializer_range (float, optional) --
The standard deviation of the normal initializer. Defaults to 0.02.
注解
A normal_initializer initializes weight matrices as normal distributions. See
BertPretrainedModel.init_weights()
for how weights are initialized inBertModel
.pad_token_id (int, optional) -- The index of padding token in the token vocabulary. Defaults to
1
.entity_pad_token_id (int, optional) -- The index of padding token in the token vocabulary. Defaults to
0
.
-
forward
(input_ids, token_type_ids=None, position_ids=None, attention_mask=None, entity_ids=None, entity_position_ids=None, entity_token_type_ids=None, entity_attention_mask=None)[源代码]¶ The LukeModel forward method, overrides the
__call__()
special method.- 参数
input_ids (Tensor) -- Indices of input sequence tokens in the vocabulary. They are numerical representations of tokens that build the input sequence. Its data type should be
int64
and it has a shape of [batch_size, sequence_length].token_type_ids (Tensor, optional) --
Segment token indices to indicate different portions of the inputs. Selected in the range
[0, type_vocab_size - 1]
. Iftype_vocab_size
is 2, which means the inputs have two portions. Indices can either be 0 or 1:0 corresponds to a sentence A token,
1 corresponds to a sentence B token.
Its data type should be
int64
and it has a shape of [batch_size, sequence_length]. Defaults toNone
, which means we don't add segment embeddings.position_ids (Tensor, optional) -- Indices of positions of each input sequence tokens in the position embeddings. Selected in the range
[0, max_position_embeddings - 1]
. Shape as(batch_size, num_tokens)
and dtype as int64. Defaults toNone
.attention_mask (Tensor, optional) -- Mask used in multi-head attention to avoid performing attention on to some unwanted positions, usually the paddings or the subsequent positions. Its data type can be int, float and bool. When the data type is bool, the
masked
tokens haveFalse
values and the others haveTrue
values. When the data type is int, themasked
tokens have0
values and the others have1
values. When the data type is float, themasked
tokens have-INF
values and the others have0
values. It is a tensor with shape broadcasted to[batch_size, num_attention_heads, sequence_length, sequence_length]
. Defaults toNone
, which means nothing needed to be prevented attention to.entity_ids (Tensor, optional) -- Indices of entity sequence tokens in the entity vocabulary. They are numerical representations of entities that build the entity input sequence. Its data type should be
int64
and it has a shape of [batch_size, entity_sequence_length].entity_position_ids (Tensor, optional) -- Indices of positions of each entity sequence tokens in the position embeddings. Selected in the range
[0, max_position_embeddings - 1]
. Shape as(batch_size, num_entity_tokens)
and dtype as int64. Defaults toNone
.entity_token_type_ids (Tensor, optional) -- Segment entity token indices to indicate different portions of the entity inputs. Selected in the range
[0, type_vocab_size - 1]
. Iftype_vocab_size
is 2, which means the inputs have two portions. Indices can either be 0 or 1:entity_attention_mask (Tensor, optional) -- Mask used in multi-head attention to avoid performing attention on to some unwanted positions, usually the paddings or the subsequent positions. Its data type can be int, float and bool. When the data type is bool, the
masked
tokens haveFalse
values and the others haveTrue
values. When the data type is int, themasked
tokens have0
values and the others have1
values. When the data type is float, themasked
tokens have-INF
values and the others have0
values. It is a tensor will be concat withattention_mask
.
- 返回
Returns tuple (
word_hidden_state, entity_hidden_state, pool_output
).With the fields:
word_hidden_state
(Tensor):Sequence of hidden-states at the last layer of the model. It's data type should be float32 and its shape is [batch_size, sequence_length, hidden_size].
entity_hidden_state
(Tensor):Sequence of entity hidden-states at the last layer of the model. It's data type should be float32 and its shape is [batch_size, sequence_length, hidden_size].
pooled_output
(Tensor):The output of first token (
<s>
) in sequence. We "pool" the model by simply taking the hidden state corresponding to the first token. Its data type should be float32 and its shape is [batch_size, hidden_size].
- 返回类型
tuple
示例
import paddle from paddlenlp.transformers import LukeModel, LukeTokenizer tokenizer = LukeTokenizer.from_pretrained('luke-base') model = LukeModel.from_pretrained('luke-base') text = "Beyoncé lives in Los Angeles." entity_spans = [(0, 7)] inputs = tokenizer(text, entity_spans=entity_spans, add_prefix_space=True) inputs = {k:paddle.to_tensor([v]) for (k, v) in inputs.items()} output = model(**inputs)
-
class
LukePretrainedModel
(*args, **kwargs)[源代码]¶ 基类:
paddlenlp.transformers.model_utils.PretrainedModel
An abstract class for pretrained Luke models. It provides Luke related
model_config_file
,pretrained_init_configuration
,resource_files_names
,pretrained_resource_files_map
,base_model_prefix
for downloading and loading pretrained models. SeePretrainedModel
for more details.-
base_model_class
¶
-
-
class
LukeForEntitySpanClassification
(luke, num_classes)[源代码]¶ 基类:
paddlenlp.transformers.luke.modeling.LukePretrainedModel
The LUKE model with a span classification head on top (a linear layer on top of the hidden states output) for tasks such as named entity recognition.
- 参数
luke (
LukeModel
) -- An instance of LukeModel.num_classes (int) -- The number of classes.
-
forward
(entity_start_positions, entity_end_positions, input_ids, token_type_ids=None, position_ids=None, attention_mask=None, entity_ids=None, entity_position_ids=None, entity_token_type_ids=None, entity_attention_mask=None)[源代码]¶ The LukeForEntitySpanClassification forward method, overrides the __call__() special method.
- 参数
entity_start_positions -- The start position of entities in sequence.
entity_end_positions -- The start position of entities in sequence.
input_ids (Tensor) -- See
LukeModel
.token_type_ids (Tensor, optional) -- See
LukeModel
.position_ids (Tensor, optional) -- See :class:
LukeModel
attention_mask (list, optional) -- See
LukeModel
.entity_ids (Tensor, optional) -- See
LukeModel
.entity_position_ids (Tensor, optional) -- See
LukeModel
.entity_token_type_ids (Tensor, optional) -- See
LukeModel
.entity_attention_mask (list, optional) -- See
LukeModel
.
- 返回
Returns tensor
logits
, a tensor of the entity span classification logits. Shape as[batch_size, num_entities, num_classes]
and dtype as float32.- 返回类型
Tensor
示例
import paddle from paddlenlp.transformers import LukeForEntitySpanClassification, LukeTokenizer tokenizer = LukeTokenizer.from_pretrained('luke-base') model = LukeForEntitySpanClassification.from_pretrained('luke-base', num_classes=2) text = "Beyoncé lives in Los Angeles." entity_spans = [(0, 7)] inputs = tokenizer(text, entity_spans=entity_spans, add_prefix_space=True) inputs = {k:paddle.to_tensor([v]) for (k, v) in inputs.items()} inputs['entity_start_positions'] = paddle.to_tensor([[1]], dtype='int64') inputs['entity_end_positions'] = paddle.to_tensor([[2]], dtype='int64') logits = model(**inputs)
-
class
LukeForEntityPairClassification
(luke, num_classes)[源代码]¶ 基类:
paddlenlp.transformers.luke.modeling.LukePretrainedModel
The LUKE model with a classification head on top (a linear layer on top of the hidden states of the two entity tokens) for entity pair classification tasks, such as TACRED.
- 参数
luke (
LukeModel
) -- An instance of LukeModel.num_classes (int) -- The number of classes.
-
forward
(input_ids, token_type_ids=None, position_ids=None, attention_mask=None, entity_ids=None, entity_position_ids=None, entity_token_type_ids=None, entity_attention_mask=None)[源代码]¶ The LukeForEntityPairClassification forward method, overrides the __call__() special method.
- 参数
input_ids (Tensor) -- See
LukeModel
.token_type_ids (Tensor, optional) -- See
LukeModel
.position_ids (Tensor, optional) -- See :class:
LukeModel
attention_mask (list, optional) -- See
LukeModel
.entity_ids (Tensor, optional) -- See
LukeModel
.entity_position_ids (Tensor, optional) -- See
LukeModel
.entity_token_type_ids (Tensor, optional) -- See
LukeModel
.entity_attention_mask (list, optional) -- See
LukeModel
.
- 返回
Returns tensor
logits
, a tensor of the entity pair classification logits. Shape as[batch_size, num_classes]
and dtype as float32.- 返回类型
Tensor
示例
import paddle from paddlenlp.transformers import LukeForEntityPairClassification, LukeTokenizer tokenizer = LukeTokenizer.from_pretrained('luke-base') model = LukeForEntityPairClassification.from_pretrained('luke-base', num_classes=2) text = "Beyoncé lives in Los Angeles." entity_spans = [(0, 7), (17, 28)] inputs = tokenizer(text, entity_spans=entity_spans, add_prefix_space=True) inputs = {k:paddle.to_tensor([v]) for (k, v) in inputs.items()} logits = model(**inputs)
-
class
LukeForEntityClassification
(luke, num_classes)[源代码]¶ 基类:
paddlenlp.transformers.luke.modeling.LukePretrainedModel
The LUKE model with a classification head on top (a linear layer on top of the hidden state of the first entity token) for entity classification tasks, such as Open Entity.
- 参数
luke (
LukeModel
) -- An instance of LukeModel.num_classes (int) -- The number of classes.
-
forward
(input_ids, token_type_ids=None, position_ids=None, attention_mask=None, entity_ids=None, entity_position_ids=None, entity_token_type_ids=None, entity_attention_mask=None)[源代码]¶ The LukeForEntityClassification forward method, overrides the __call__() special method.
- 参数
input_ids (Tensor) -- See
LukeModel
.token_type_ids (Tensor, optional) -- See
LukeModel
.position_ids (Tensor, optional) -- See :class:
LukeModel
attention_mask (list, optional) -- See
LukeModel
.entity_ids (Tensor, optional) -- See
LukeModel
.entity_position_ids (Tensor, optional) -- See
LukeModel
.entity_token_type_ids (Tensor, optional) -- See
LukeModel
.entity_attention_mask (list, optional) -- See
LukeModel
.
- 返回
Returns tensor
logits
, a tensor of the entity classification logits. Shape as[batch_size, num_classes]
and dtype as float32.- 返回类型
Tensor
示例
import paddle from paddlenlp.transformers import LukeForEntityClassification, LukeTokenizer tokenizer = LukeTokenizer.from_pretrained('luke-base') model = LukeForEntityClassification.from_pretrained('luke-base', num_classes=2) text = "Beyoncé lives in Los Angeles." entity_spans = [(0, 7)] inputs = tokenizer(text, entity_spans=entity_spans, add_prefix_space=True) inputs = {k:paddle.to_tensor([v]) for (k, v) in inputs.items()} logits = model(**inputs)
-
class
LukeForMaskedLM
(luke)[源代码]¶ 基类:
paddlenlp.transformers.luke.modeling.LukePretrainedModel
Luke Model with a
masked language modeling
head on top.-
forward
(input_ids, token_type_ids=None, position_ids=None, attention_mask=None, entity_ids=None, entity_position_ids=None, entity_token_type_ids=None, entity_attention_mask=None)[源代码]¶ The LukeForMaskedLM forward method, overrides the __call__() special method.
- 参数
input_ids (Tensor) -- See
LukeModel
.token_type_ids (Tensor, optional) -- See
LukeModel
.position_ids (Tensor, optional) -- See :class:
LukeModel
attention_mask (list, optional) -- See
LukeModel
.entity_ids (Tensor, optional) -- See
LukeModel
.entity_position_ids (Tensor, optional) -- See
LukeModel
.entity_token_type_ids (Tensor, optional) -- See
LukeModel
.entity_attention_mask (list, optional) -- See
LukeModel
.
- 返回
Returns tuple (
logits
,entity_logits
).With the fields:
logits
(Tensor):The scores of masked token prediction. Its data type should be float32 and shape is [batch_size, sequence_length, vocab_size].
entity_logits
(Tensor):The scores of masked entity prediction. Its data type should be float32 and its shape is [batch_size, entity_length, entity_vocab_size].
- 返回类型
tuple
示例
import paddle from paddlenlp.transformers import LukeForMaskedLM, LukeTokenizer tokenizer = LukeTokenizer.from_pretrained('luke-base') model = LukeForMaskedLM.from_pretrained('luke-base') text = "Beyoncé lives in Los Angeles." entity_spans = [(0, 7)] inputs = tokenizer(text, entity_spans=entity_spans, add_prefix_space=True) inputs = {k:paddle.to_tensor([v]) for (k, v) in inputs.items()} logits, entity_logits = model(**inputs)
-
-
class
LukeForQuestionAnswering
(luke)[源代码]¶ 基类:
paddlenlp.transformers.luke.modeling.LukePretrainedModel
LukeBert Model with question answering tasks. :param luke: An instance of
LukeModel
. :type luke:LukeModel
-
forward
(input_ids=None, token_type_ids=None, position_ids=None, attention_mask=None, entity_ids=None, entity_position_ids=None, entity_token_type_ids=None, entity_attention_mask=None)[源代码]¶ The LukeForQuestionAnswering forward method, overrides the __call__() special method.
- 参数
input_ids (Tensor) -- See
LukeModel
.token_type_ids (Tensor, optional) -- See
LukeModel
.position_ids (Tensor, optional) -- See :class:
LukeModel
attention_mask (list, optional) -- See
LukeModel
.entity_ids (Tensor, optional) -- See
LukeModel
.entity_position_ids (Tensor, optional) -- See
LukeModel
.entity_token_type_ids (Tensor, optional) -- See
LukeModel
.entity_attention_mask (list, optional) -- See
LukeModel
.
- 返回
Returns tuple (
start_logits
,end_logits
). With the fields: -start_logits
(Tensor):A tensor of the input token classification logits, indicates the start position of the labelled span. Its data type should be float32 and its shape is [batch_size, sequence_length].
end_logits
(Tensor):A tensor of the input token classification logits, indicates the end position of the labelled span. Its data type should be float32 and its shape is [batch_size, sequence_length].
- 返回类型
tuple
示例
import paddle from paddlenlp.transformers import LukeForQuestionAnswering, LukeTokenizer tokenizer = LukeTokenizer.from_pretrained('luke-base') model = LukeForQuestionAnswering.from_pretrained('luke-base') text = "Beyoncé lives in Los Angeles." entity_spans = [(0, 7)] inputs = tokenizer(text, entity_spans=entity_spans, add_prefix_space=True) inputs = {k:paddle.to_tensor([v]) for (k, v) in inputs.items()} start_logits, end_logits = model(**inputs)
-