haipradana/indobert-indonesia-satire-sarcastic-classification-model

Fine-tuned indoBERT pre-trained model for sarcasm and satire classification

Just check GitHub for full-code and Google Colab: https://github.com/haipradana/indobert-indonesia-sarcastic-satire-classification

How to use this model?

import torch
from transformers import AutoTokenizer, AutoModelForSequenceClassification

# Load model
tokenizer = AutoTokenizer.from_pretrained("indobert-indonesia-sarcastic-satire-classification/model")
model = AutoModelForSequenceClassification.from_pretrained("indobert-indonesia-sarcastic-satire-classification/model")

# Predict
def predict(text: str):
    inputs = tokenizer(text, return_tensors='pt', truncation=True, padding=True, max_length=512)
    with torch.no_grad():
        outputs = model(**inputs)
    logits = outputs.logits
    prediction = torch.argmax(logits, dim=1).item()
    return 'sarcasm' if prediction == 1 else 'not sarcasm'

# Example
result = predict("Kamu penulis ya? pandai sekali mengarang cerita")
print(result) #output = sarcasm

Or just using the script in the GitHub Repos

cd scripts
python predict.py

Evaluation Results

The model was evaluated on the test set with the following metrics:

Metric	Value
Accuracy	0.8378
Precision	0.8405
Recall	0.8286
F1-Score	0.8345

Training History

Epoch	Train Loss	Val Loss	Accuracy	Precision	F1-Score	Recall
1	0.4559	0.3512	0.8409	0.9022	0.8261	0.7618
2	0.2491	0.3924	0.8339	0.7835	0.8459	0.9190
3	0.1198	0.5980	0.8429	0.8188	0.8471	0.8774
4	0.0439	0.9497	0.8444	0.8231	0.8479	0.8742
5	0.0097	0.9962	0.8522	0.8421	0.8529	0.8640