File size: 3,983 Bytes
0fde84f | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 | import torch
import torch.nn as nn
from torch.nn import CrossEntropyLoss, MSELoss
from transformers.activations import get_activation
from transformers import (
ElectraPreTrainedModel,
ElectraModel,
ElectraConfig,
ElectraTokenizer,
BertConfig,
BertTokenizer
)
# MODEL_CLASSES = {
# "koelectra-base": (ElectraConfig, koElectraForSequenceClassification, ElectraTokenizer),
# "koelectra-small": (ElectraConfig, koElectraForSequenceClassification, ElectraTokenizer),
# "koelectra-base-v2": (ElectraConfig, koElectraForSequenceClassification, ElectraTokenizer),
# "koelectra-small-v2": (ElectraConfig, koElectraForSequenceClassification, ElectraTokenizer),
# }
# def load_tokenizer(args):
# return MODEL_CLASSES[args.model_type][2].from_pretrained(args.model_name_or_path)
class ElectraClassificationHead(nn.Module):
"""Head for sentence-level classification tasks."""
def __init__(self, config, num_labels):
super().__init__()
self.dense = nn.Linear(config.hidden_size, 4*config.hidden_size)
self.dropout = nn.Dropout(config.hidden_dropout_prob)
self.out_proj = nn.Linear(4*config.hidden_size,num_labels)
def forward(self, features, **kwargs):
x = features[:, 0, :] # take <s> token (equiv. to [CLS])
x = self.dropout(x)
x = self.dense(x)
x = get_activation("gelu")(x) # although BERT uses tanh here, it seems Electra authors used gelu here
x = self.dropout(x)
x = self.out_proj(x)
return x
class koElectraForSequenceClassification(ElectraPreTrainedModel):
def __init__(self,
config,
num_labels):
super().__init__(config)
self.num_labels = num_labels
self.electra = ElectraModel(config)
self.classifier = ElectraClassificationHead(config, num_labels)
self.init_weights()
def forward(
self,
input_ids=None,
attention_mask=None,
token_type_ids=None,
position_ids=None,
head_mask=None,
inputs_embeds=None,
labels=None,
output_attentions=None,
output_hidden_states=None,
):
r"""
labels (:obj:`torch.LongTensor` of shape :obj:`(batch_size,)`, `optional`, defaults to :obj:`None`):
Labels for computing the sequence classification/regression loss.
Indices should be in :obj:`[0, ..., config.num_labels - 1]`.
If :obj:`config.num_labels == 1` a regression loss is computed (Mean-Square loss),
If :obj:`config.num_labels > 1` a classification loss is computed (Cross-Entropy).
"""
discriminator_hidden_states = self.electra(
input_ids,
attention_mask,
token_type_ids,
position_ids,
head_mask,
inputs_embeds,
output_attentions,
output_hidden_states,
)
sequence_output = discriminator_hidden_states[0]
logits = self.classifier(sequence_output)
outputs = (logits,) + discriminator_hidden_states[1:] # add hidden states and attention if they are here
if labels is not None:
if self.num_labels == 1:
# We are doing regression
loss_fct = MSELoss()
loss = loss_fct(logits.view(-1), labels.view(-1))
else:
loss_fct = CrossEntropyLoss()
loss = loss_fct(logits.view(-1, self.num_labels), labels.view(-1))
outputs = (loss,) + outputs
return outputs # (loss), (logits), (hidden_states), (attentions)
def koelectra_input(tokenizer, str, device = None, max_seq_len = 512):
index_of_words = tokenizer.encode(str)
# token_type_ids = [0] * len(index_of_words)
attention_mask = [1] * len(index_of_words)
# Padding Length
padding_length = max_seq_len - len(index_of_words)
# Zero Padding
index_of_words += [0] * padding_length
# token_type_ids += [0] * padding_length
attention_mask += [0] * padding_length
data = {
'input_ids': torch.tensor([index_of_words]).to(device),
'attention_mask': torch.tensor([attention_mask]).to(device),
}
return data |