model.py

# Nama file: model.py

import torch
import torch.nn as nn
from transformers import AutoModel, AutoConfig

class IndoBERTClassifier(nn.Module):
    def __init__(self, config):
        super(IndoBERTClassifier, self).__init__()
        # Gunakan config dari model dasar untuk mengambil hidden_size
        self.bert = AutoModel.from_pretrained(config._name_or_path, config=config)
        self.dropout = nn.Dropout(config.classifier_dropout if hasattr(config, 'classifier_dropout') else 0.1)
        hidden_size = self.bert.config.hidden_size

        self.num_clickbait_labels = config.num_clickbait_labels
        self.num_kategori_labels = config.num_kategori_labels

        self.clickbait_classifier = nn.Linear(hidden_size, self.num_clickbait_labels)
        self.kategori_classifier = nn.Linear(hidden_size, self.num_kategori_labels)

    def forward(self, input_ids, attention_mask, clickbait_labels=None, kategori_labels=None, **kwargs):
        output = self.bert(input_ids=input_ids, attention_mask=attention_mask)
        pooled_output = output.last_hidden_state[:, 0, :]  # Ambil token [CLS]

        dropout_output = self.dropout(pooled_output)

        clickbait_logits = self.clickbait_classifier(dropout_output)
        kategori_logits = self.kategori_classifier(dropout_output)

        return {
            "clickbait_logits": clickbait_logits,
            "kategori_logits": kategori_logits
        }