Spaces:
Sleeping
Sleeping
import torch | |
from torch import nn | |
from transformers import AutoTokenizer, AutoModelForSequenceClassification | |
class BERTClassifier(nn.Module): | |
def __init__(self): | |
super().__init__() | |
self.bert = AutoModelForSequenceClassification.from_pretrained('cointegrated/rubert-tiny-toxicity') | |
self.bert.classifier = nn.Linear(312, 312) | |
for param in self.bert.parameters(): | |
param.requires_grad = False | |
self.linear = nn.Sequential( | |
nn.Linear(312, 128), | |
nn.Sigmoid(), | |
nn.Dropout(), | |
nn.Linear(128, 1) | |
) | |
def forward(self, x, attention_mask=None): | |
bert_out = self.bert(x, attention_mask=attention_mask).logits | |
out = self.linear(bert_out).squeeze(1) | |
return out |