SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

[inputs_embeds] All PyTorch models

Browse Source
This commit is contained in:
Julien Chaumond
2019-11-05 00:39:18 +00:00
parent 9eddf44b7a
commit 00337e9687
21 changed files with 361 additions and 147 deletions
Download Patch File Download Diff File Expand all files Collapse all files

47
transformers/modeling_roberta.py
View File

@@ -48,16 +48,24 @@ class RobertaEmbeddings(BertEmbeddings):
self.position_embeddings = nn.Embedding(config.max_position_embeddings, config.hidden_size,
padding_idx=self.padding_idx)
def forward(self, input_ids, token_type_ids=None, position_ids=None):
seq_length = input_ids.size(1)
def forward(self, input_ids=None, token_type_ids=None, position_ids=None, inputs_embeds=None):
if input_ids is not None:
input_shape = input_ids.size()
else:
input_shape = inputs_embeds.size()[:-1]
seq_length = input_shape[1]
device = input_ids.device if input_ids is not None else inputs_embeds.device
if position_ids is None:
# Position numbers begin at padding_idx+1. Padding symbols are ignored.
# cf. fairseq's `utils.make_positions`
position_ids = torch.arange(self.padding_idx+1, seq_length+self.padding_idx+1, dtype=torch.long, device=input_ids.device)
position_ids = position_ids.unsqueeze(0).expand_as(input_ids)
position_ids = torch.arange(self.padding_idx+1, seq_length+self.padding_idx+1, dtype=torch.long, device=device)
position_ids = position_ids.unsqueeze(0).expand(input_shape)
return super(RobertaEmbeddings, self).forward(input_ids,
token_type_ids=token_type_ids,
position_ids=position_ids)
position_ids=position_ids,
inputs_embeds=inputs_embeds)
ROBERTA_START_DOCSTRING = r""" The RoBERTa model was proposed in
@@ -126,6 +134,10 @@ ROBERTA_INPUTS_DOCSTRING = r"""
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**: (`optional`) ``torch.FloatTensor`` of shape ``(batch_size, sequence_length, embedding_dim)``:
Optionally, instead of passing ``input_ids`` you can choose to directly pass an embedded representation.
This is useful if you want more control over how to convert `input_ids` indices into associated vectors
than the model's internal embedding lookup matrix.
"""
@add_start_docstrings("The bare RoBERTa Model transformer outputting raw hidden-states without any specific head on top.",
@@ -222,13 +234,14 @@ class RobertaForMaskedLM(BertPreTrainedModel):
def get_output_embeddings(self):
return self.lm_head.decoder
def forward(self, input_ids, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None,
def forward(self, input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None,
masked_lm_labels=None):
outputs = self.roberta(input_ids,
attention_mask=attention_mask,
token_type_ids=token_type_ids,
position_ids=position_ids,
head_mask=head_mask)
head_mask=head_mask,
inputs_embeds=inputs_embeds)
sequence_output = outputs[0]
prediction_scores = self.lm_head(sequence_output)
@@ -309,13 +322,14 @@ class RobertaForSequenceClassification(BertPreTrainedModel):
self.roberta = RobertaModel(config)
self.classifier = RobertaClassificationHead(config)
def forward(self, input_ids, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None,
def forward(self, input_ids=None, attention_mask=None, token_type_ids=None, position_ids=None, head_mask=None, inputs_embeds=None,
labels=None):
outputs = self.roberta(input_ids,
attention_mask=attention_mask,
token_type_ids=token_type_ids,
position_ids=position_ids,
head_mask=head_mask)
head_mask=head_mask,
inputs_embeds=inputs_embeds)
sequence_output = outputs[0]
logits = self.classifier(sequence_output)
@@ -372,6 +386,10 @@ class RobertaForMultipleChoice(BertPreTrainedModel):
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**: (`optional`) ``torch.FloatTensor`` of shape ``(batch_size, sequence_length, embedding_dim)``:
Optionally, instead of passing ``input_ids`` you can choose to directly pass an embedded representation.
This is useful if you want more control over how to convert `input_ids` indices into associated vectors
than the model's internal embedding lookup matrix.
**labels**: (`optional`) ``torch.LongTensor`` of shape ``(batch_size,)``:
Labels for computing the multiple choice classification loss.
Indices should be in ``[0, ..., num_choices]`` where `num_choices` is the size of the second dimension
@@ -415,8 +433,8 @@ class RobertaForMultipleChoice(BertPreTrainedModel):
self.init_weights()
def forward(self, input_ids, token_type_ids=None, attention_mask=None, labels=None,
position_ids=None, head_mask=None):
def forward(self, input_ids=None, token_type_ids=None, attention_mask=None, labels=None,
position_ids=None, head_mask=None, inputs_embeds=None):
num_choices = input_ids.shape[1]
flat_input_ids = input_ids.view(-1, input_ids.size(-1))
@@ -487,14 +505,15 @@ class RobertaForTokenClassification(BertPreTrainedModel):
self.init_weights()
def forward(self, input_ids, attention_mask=None, token_type_ids=None,
position_ids=None, head_mask=None, labels=None):
def forward(self, input_ids=None, attention_mask=None, token_type_ids=None,
position_ids=None, head_mask=None, inputs_embeds=None, labels=None):
outputs = self.roberta(input_ids,
attention_mask=attention_mask,
token_type_ids=token_type_ids,
position_ids=position_ids,
head_mask=head_mask)
head_mask=head_mask,
inputs_embeds=inputs_embeds)
sequence_output = outputs[0]