Spaces:
Sleeping
Sleeping
File size: 4,019 Bytes
faa8cd6 a91e942 faa8cd6 a91e942 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 |
import torch
import torch.nn as nn
from transformers.models.roberta.modeling_roberta import *
class MRCQuestionAnswering(RobertaPreTrainedModel):
config_class = RobertaConfig
def _reorder_cache(self, past, beam_idx):
pass
_keys_to_ignore_on_load_unexpected = [r"pooler"]
_keys_to_ignore_on_load_missing = [r"position_ids"]
def __init__(self, config):
super().__init__(config)
self.num_labels = config.num_labels
self.roberta = RobertaModel(config, add_pooling_layer=False)
self.qa_outputs = nn.Linear(config.hidden_size, config.num_labels)
self.init_weights()
def forward(
self,
input_ids=None,
words_lengths=None,
start_idx=None,
end_idx=None,
attention_mask=None,
token_type_ids=None,
position_ids=None,
head_mask=None,
inputs_embeds=None,
start_positions=None,
end_positions=None,
span_answer_ids=None,
output_attentions=None,
output_hidden_states=None,
return_dict=None,
):
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
outputs = self.roberta(
input_ids,
attention_mask=attention_mask,
token_type_ids=token_type_ids,
position_ids=position_ids,
head_mask=head_mask,
inputs_embeds=inputs_embeds,
output_attentions=output_attentions,
output_hidden_states=output_hidden_states,
return_dict=return_dict,
)
sequence_output = outputs[0]
context_embedding = sequence_output
# Compute align word sub_word matrix
batch_size = input_ids.shape[0]
max_sub_word = input_ids.shape[1]
max_word = words_lengths.shape[1]
align_matrix = torch.zeros((batch_size, max_word, max_sub_word))
for i, sample_length in enumerate(words_lengths):
for j in range(len(sample_length)):
start_idx = torch.sum(sample_length[:j])
align_matrix[i][j][start_idx: start_idx + sample_length[j]] = 1 if sample_length[j] > 0 else 0
align_matrix = align_matrix.to(context_embedding.device)
# Combine sub_word features to make word feature
context_embedding_align = torch.bmm(align_matrix, context_embedding)
logits = self.qa_outputs(context_embedding_align)
start_logits, end_logits = logits.split(1, dim=-1)
start_logits = start_logits.squeeze(-1).contiguous()
end_logits = end_logits.squeeze(-1).contiguous()
total_loss = None
if start_positions is not None and end_positions is not None:
# If we are on multi-GPU, split add a dimension
if len(start_positions.size()) > 1:
start_positions = start_positions.squeeze(-1)
if len(end_positions.size()) > 1:
end_positions = end_positions.squeeze(-1)
# sometimes the start/end positions are outside our model inputs, we ignore these terms
ignored_index = start_logits.size(1)
start_positions = start_positions.clamp(0, ignored_index)
end_positions = end_positions.clamp(0, ignored_index)
loss_fct = nn.CrossEntropyLoss(ignore_index=ignored_index)
start_loss = loss_fct(start_logits, start_positions)
end_loss = loss_fct(end_logits, end_positions)
total_loss = (start_loss + end_loss) / 2
if not return_dict:
output = (start_logits, end_logits) + outputs[2:]
return ((total_loss,) + output) if total_loss is not None else output
return QuestionAnsweringModelOutput(
loss=total_loss,
start_logits=start_logits,
end_logits=end_logits,
hidden_states=outputs.hidden_states,
attentions=outputs.attentions,
) |