Skip to content

GitLab

"...git@developer.sourcefind.cn:chenpangpang/transformers.git" did not exist on "53d891247bd45f316d3d710c64b54964bea14a64"
Unverified Commit 41bfc1e2 authored by Zehua Li's avatar Zehua Li Committed by GitHub
Browse files

Add type hints for ConvBert model (#16377) 



* Add missing type hints for ConvBERT flavored models.

* Update src/transformers/models/convbert/modeling_convbert.py
Co-authored-by: default avatarMatt <Rocketknight1@users.noreply.github.com>
parent 23a75a53
Hide whitespace changes
Inline Side-by-side
Showing with 122 additions and 117 deletions
src/transformers/models/convbert/modeling_convbert.py
View file @ 41bfc1e2
...@@ -18,6 +18,7 @@ ...@@ -18,6 +18,7 @@
import math import math
import os import os
from operator import attrgetter from operator import attrgetter
from typing import Optional, Tuple, Union
import torch import torch
import torch.utils.checkpoint import torch.utils.checkpoint
...@@ -902,17 +903,17 @@ class ConvBertForMaskedLM(ConvBertPreTrainedModel): ...@@ -902,17 +903,17 @@ class ConvBertForMaskedLM(ConvBertPreTrainedModel):
) )
def forward( def forward(
self, self,
input_ids=None, input_ids: Optional[torch.LongTensor] = None,
attention_mask=None, attention_mask: Optional[torch.FloatTensor] = None,
token_type_ids=None, token_type_ids: Optional[torch.LongTensor] = None,
position_ids=None, position_ids: Optional[torch.LongTensor] = None,
head_mask=None, head_mask: Optional[torch.FloatTensor] = None,
inputs_embeds=None, inputs_embeds: Optional[torch.FloatTensor] = None,
labels=None, labels: Optional[torch.LongTensor] = None,
output_attentions=None, output_attentions: Optional[bool] = None,
output_hidden_states=None, output_hidden_states: Optional[bool] = None,
return_dict=None, return_dict: Optional[bool] = None,
): ) -> Union[Tuple, MaskedLMOutput]:
r""" r"""
labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*): labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ..., Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ...,
...@@ -1006,17 +1007,17 @@ class ConvBertForSequenceClassification(ConvBertPreTrainedModel): ...@@ -1006,17 +1007,17 @@ class ConvBertForSequenceClassification(ConvBertPreTrainedModel):
) )
def forward( def forward(
self, self,
input_ids=None, input_ids: Optional[torch.LongTensor] = None,
attention_mask=None, attention_mask: Optional[torch.FloatTensor] = None,
token_type_ids=None, token_type_ids: Optional[torch.LongTensor] = None,
position_ids=None, position_ids: Optional[torch.LongTensor] = None,
head_mask=None, head_mask: Optional[torch.FloatTensor] = None,
inputs_embeds=None, inputs_embeds: Optional[torch.FloatTensor] = None,
labels=None, labels: Optional[torch.LongTensor] = None,
output_attentions=None, output_attentions: Optional[bool] = None,
output_hidden_states=None, output_hidden_states: Optional[bool] = None,
return_dict=None, return_dict: Optional[bool] = None,
): ) -> Union[Tuple, SequenceClassifierOutput]:
r""" r"""
labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*): labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
Labels for computing the sequence classification/regression loss. Indices should be in `[0, ..., Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,
...@@ -1104,17 +1105,17 @@ class ConvBertForMultipleChoice(ConvBertPreTrainedModel): ...@@ -1104,17 +1105,17 @@ class ConvBertForMultipleChoice(ConvBertPreTrainedModel):
) )
def forward( def forward(
self, self,
input_ids=None, input_ids: Optional[torch.LongTensor] = None,
attention_mask=None, attention_mask: Optional[torch.FloatTensor] = None,
token_type_ids=None, token_type_ids: Optional[torch.LongTensor] = None,
position_ids=None, position_ids: Optional[torch.LongTensor] = None,
head_mask=None, head_mask: Optional[torch.FloatTensor] = None,
inputs_embeds=None, inputs_embeds: Optional[torch.FloatTensor] = None,
labels=None, labels: Optional[torch.LongTensor] = None,
output_attentions=None, output_attentions: Optional[bool] = None,
output_hidden_states=None, output_hidden_states: Optional[bool] = None,
return_dict=None, return_dict: Optional[bool] = None,
): ) -> Union[Tuple, MultipleChoiceModelOutput]:
r""" r"""
labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*): labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
Labels for computing the multiple choice classification loss. Indices should be in `[0, ..., Labels for computing the multiple choice classification loss. Indices should be in `[0, ...,
...@@ -1200,17 +1201,17 @@ class ConvBertForTokenClassification(ConvBertPreTrainedModel): ...@@ -1200,17 +1201,17 @@ class ConvBertForTokenClassification(ConvBertPreTrainedModel):
) )
def forward( def forward(
self, self,
input_ids=None, input_ids: Optional[torch.LongTensor] = None,
attention_mask=None, attention_mask: Optional[torch.FloatTensor] = None,
token_type_ids=None, token_type_ids: Optional[torch.LongTensor] = None,
position_ids=None, position_ids: Optional[torch.LongTensor] = None,
head_mask=None, head_mask: Optional[torch.FloatTensor] = None,
inputs_embeds=None, inputs_embeds: Optional[torch.FloatTensor] = None,
labels=None, labels: Optional[torch.LongTensor] = None,
output_attentions=None, output_attentions: Optional[bool] = None,
output_hidden_states=None, output_hidden_states: Optional[bool] = None,
return_dict=None, return_dict: Optional[bool] = None,
): ) -> Union[Tuple, TokenClassifierOutput]:
r""" r"""
labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*): labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
Labels for computing the token classification loss. Indices should be in `[0, ..., config.num_labels - 1]`. Labels for computing the token classification loss. Indices should be in `[0, ..., config.num_labels - 1]`.
...@@ -1278,18 +1279,18 @@ class ConvBertForQuestionAnswering(ConvBertPreTrainedModel): ...@@ -1278,18 +1279,18 @@ class ConvBertForQuestionAnswering(ConvBertPreTrainedModel):
) )
def forward( def forward(
self, self,
input_ids=None, input_ids: Optional[torch.LongTensor] = None,
attention_mask=None, attention_mask: Optional[torch.FloatTensor] = None,
token_type_ids=None, token_type_ids: Optional[torch.LongTensor] = None,
position_ids=None, position_ids: Optional[torch.LongTensor] = None,
head_mask=None, head_mask: Optional[torch.FloatTensor] = None,
inputs_embeds=None, inputs_embeds: Optional[torch.FloatTensor] = None,
start_positions=None, start_positions: Optional[torch.LongTensor] = None,
end_positions=None, end_positions: Optional[torch.LongTensor] = None,
output_attentions=None, output_attentions: Optional[bool] = None,
output_hidden_states=None, output_hidden_states: Optional[bool] = None,
return_dict=None, return_dict: Optional[bool] = None,
): ) -> Union[Tuple, QuestionAnsweringModelOutput]:
r""" r"""
start_positions (`torch.LongTensor` of shape `(batch_size,)`, *optional*): start_positions (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
Labels for position (index) of the start of the labelled span for computing the token classification loss. Labels for position (index) of the start of the labelled span for computing the token classification loss.
......
src/transformers/models/convbert/modeling_tf_convbert.py
View file @ 41bfc1e2
...@@ -15,6 +15,9 @@ ...@@ -15,6 +15,9 @@
""" TF 2.0 ConvBERT model.""" """ TF 2.0 ConvBERT model."""
from typing import Optional, Tuple, Union
import numpy as np
import tensorflow as tf import tensorflow as tf
from ...activations_tf import get_tf_activation from ...activations_tf import get_tf_activation
...@@ -28,6 +31,7 @@ from ...modeling_tf_outputs import ( ...@@ -28,6 +31,7 @@ from ...modeling_tf_outputs import (
) )
from ...modeling_tf_utils import ( from ...modeling_tf_utils import (
TFMaskedLanguageModelingLoss, TFMaskedLanguageModelingLoss,
TFModelInputType,
TFMultipleChoiceLoss, TFMultipleChoiceLoss,
TFPreTrainedModel, TFPreTrainedModel,
TFQuestionAnsweringLoss, TFQuestionAnsweringLoss,
...@@ -891,19 +895,19 @@ class TFConvBertForMaskedLM(TFConvBertPreTrainedModel, TFMaskedLanguageModelingL ...@@ -891,19 +895,19 @@ class TFConvBertForMaskedLM(TFConvBertPreTrainedModel, TFMaskedLanguageModelingL
) )
def call( def call(
self, self,
input_ids=None, input_ids: Optional[TFModelInputType] = None,
attention_mask=None, attention_mask: Optional[Union[np.ndarray, tf.Tensor]] = None,
token_type_ids=None, token_type_ids: Optional[Union[np.ndarray, tf.Tensor]] = None,
position_ids=None, position_ids: Optional[Union[np.ndarray, tf.Tensor]] = None,
head_mask=None, head_mask: Optional[Union[np.ndarray, tf.Tensor]] = None,
inputs_embeds=None, inputs_embeds: Optional[tf.Tensor] = None,
output_attentions=None, output_attentions: Optional[bool] = None,
output_hidden_states=None, output_hidden_states: Optional[bool] = None,
return_dict=None, return_dict: Optional[bool] = None,
labels=None, labels: Optional[tf.Tensor] = None,
training=False, training: Optional[bool] = False,
**kwargs, **kwargs,
): ) -> Union[Tuple, TFMaskedLMOutput]:
r""" r"""
labels (`tf.Tensor` of shape `(batch_size, sequence_length)`, *optional*): labels (`tf.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ..., Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ...,
...@@ -1015,19 +1019,19 @@ class TFConvBertForSequenceClassification(TFConvBertPreTrainedModel, TFSequenceC ...@@ -1015,19 +1019,19 @@ class TFConvBertForSequenceClassification(TFConvBertPreTrainedModel, TFSequenceC
) )
def call( def call(
self, self,
input_ids=None, input_ids: Optional[TFModelInputType] = None,
attention_mask=None, attention_mask: Optional[Union[np.ndarray, tf.Tensor]] = None,
token_type_ids=None, token_type_ids: Optional[Union[np.ndarray, tf.Tensor]] = None,
position_ids=None, position_ids: Optional[Union[np.ndarray, tf.Tensor]] = None,
head_mask=None, head_mask: Optional[Union[np.ndarray, tf.Tensor]] = None,
inputs_embeds=None, inputs_embeds: Optional[tf.Tensor] = None,
output_attentions=None, output_attentions: Optional[bool] = None,
output_hidden_states=None, output_hidden_states: Optional[bool] = None,
return_dict=None, return_dict: Optional[bool] = None,
labels=None, labels: Optional[tf.Tensor] = None,
training=False, training: Optional[bool] = False,
**kwargs, **kwargs,
): ) -> Union[Tuple, TFSequenceClassifierOutput]:
r""" r"""
labels (`tf.Tensor` of shape `(batch_size,)`, *optional*): labels (`tf.Tensor` of shape `(batch_size,)`, *optional*):
Labels for computing the sequence classification/regression loss. Indices should be in `[0, ..., Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,
...@@ -1124,19 +1128,19 @@ class TFConvBertForMultipleChoice(TFConvBertPreTrainedModel, TFMultipleChoiceLos ...@@ -1124,19 +1128,19 @@ class TFConvBertForMultipleChoice(TFConvBertPreTrainedModel, TFMultipleChoiceLos
) )
def call( def call(
self, self,
input_ids=None, input_ids: Optional[TFModelInputType] = None,
attention_mask=None, attention_mask: Optional[Union[np.ndarray, tf.Tensor]] = None,
token_type_ids=None, token_type_ids: Optional[Union[np.ndarray, tf.Tensor]] = None,
position_ids=None, position_ids: Optional[Union[np.ndarray, tf.Tensor]] = None,
head_mask=None, head_mask: Optional[Union[np.ndarray, tf.Tensor]] = None,
inputs_embeds=None, inputs_embeds: Optional[tf.Tensor] = None,
output_attentions=None, output_attentions: Optional[bool] = None,
output_hidden_states=None, output_hidden_states: Optional[bool] = None,
return_dict=None, return_dict: Optional[bool] = None,
labels=None, labels: Optional[tf.Tensor] = None,
training=False, training: Optional[bool] = False,
**kwargs, **kwargs,
): ) -> Union[Tuple, TFMultipleChoiceModelOutput]:
r""" r"""
labels (`tf.Tensor` of shape `(batch_size,)`, *optional*): labels (`tf.Tensor` of shape `(batch_size,)`, *optional*):
Labels for computing the multiple choice classification loss. Indices should be in `[0, ..., num_choices]` Labels for computing the multiple choice classification loss. Indices should be in `[0, ..., num_choices]`
...@@ -1261,19 +1265,19 @@ class TFConvBertForTokenClassification(TFConvBertPreTrainedModel, TFTokenClassif ...@@ -1261,19 +1265,19 @@ class TFConvBertForTokenClassification(TFConvBertPreTrainedModel, TFTokenClassif
) )
def call( def call(
self, self,
input_ids=None, input_ids: Optional[TFModelInputType] = None,
attention_mask=None, attention_mask: Optional[Union[np.ndarray, tf.Tensor]] = None,
token_type_ids=None, token_type_ids: Optional[Union[np.ndarray, tf.Tensor]] = None,
position_ids=None, position_ids: Optional[Union[np.ndarray, tf.Tensor]] = None,
head_mask=None, head_mask: Optional[Union[np.ndarray, tf.Tensor]] = None,
inputs_embeds=None, inputs_embeds: Optional[tf.Tensor] = None,
output_attentions=None, output_attentions: Optional[bool] = None,
output_hidden_states=None, output_hidden_states: Optional[bool] = None,
return_dict=None, return_dict: Optional[bool] = None,
labels=None, labels: Optional[tf.Tensor] = None,
training=False, training: Optional[bool] = False,
**kwargs, **kwargs,
): ) -> Union[Tuple, TFTokenClassifierOutput]:
r""" r"""
labels (`tf.Tensor` of shape `(batch_size, sequence_length)`, *optional*): labels (`tf.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
Labels for computing the token classification loss. Indices should be in `[0, ..., config.num_labels - 1]`. Labels for computing the token classification loss. Indices should be in `[0, ..., config.num_labels - 1]`.
...@@ -1355,20 +1359,20 @@ class TFConvBertForQuestionAnswering(TFConvBertPreTrainedModel, TFQuestionAnswer ...@@ -1355,20 +1359,20 @@ class TFConvBertForQuestionAnswering(TFConvBertPreTrainedModel, TFQuestionAnswer
) )
def call( def call(
self, self,
input_ids=None, input_ids: Optional[TFModelInputType] = None,
attention_mask=None, attention_mask: Optional[Union[np.ndarray, tf.Tensor]] = None,
token_type_ids=None, token_type_ids: Optional[Union[np.ndarray, tf.Tensor]] = None,
position_ids=None, position_ids: Optional[Union[np.ndarray, tf.Tensor]] = None,
head_mask=None, head_mask: Optional[Union[np.ndarray, tf.Tensor]] = None,
inputs_embeds=None, inputs_embeds: Optional[tf.Tensor] = None,
output_attentions=None, output_attentions: Optional[bool] = None,
output_hidden_states=None, output_hidden_states: Optional[bool] = None,
return_dict=None, return_dict: Optional[bool] = None,
start_positions=None, start_positions: Optional[tf.Tensor] = None,
end_positions=None, end_positions: Optional[tf.Tensor] = None,
training=False, training: Optional[bool] = False,
**kwargs, **kwargs,
): ) -> Union[Tuple, TFQuestionAnsweringModelOutput]:
r""" r"""
start_positions (`tf.Tensor` of shape `(batch_size,)`, *optional*): start_positions (`tf.Tensor` of shape `(batch_size,)`, *optional*):
Labels for position (index) of the start of the labelled span for computing the token classification loss. Labels for position (index) of the start of the labelled span for computing the token classification loss.
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment