Downloads · 30 days
5
8% of all-time downloads
nmt-gaul-formal
SateCincau/nmt-gaul-formal
nmt-gaul-formal is a machine learning model from SateCincau. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as mit.
Model Sequence-to-Sequence berbasis Transformer untuk mengonversi teks informal dari ulasan e-commerce ke bahasa Indonesia formal.
All-time downloads
59
Public
Parameters
60.5M
484 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors242 MB · 99%
At a glance
- License
- mit
- Model type
- t5
- Access
- Public
- Created
- Aug 25, 2026
- Updated
- Sep 19, 2026
- SHA
- d0a7c795
Datasets
Model card
From the Hugging Face model README
NMT Text Formalization E-Commerce Indonesia (Fine-tuned T5)
Model Sequence-to-Sequence berbasis Transformer untuk mengonversi teks informal dari ulasan e-commerce ke bahasa Indonesia formal.
Informasi Model
- Base Model: panggi/t5-small-indonesian-summarization-cased
- Arsitektur: T5 Sequence-to-Sequence (Transformer)
- Task: Monolingual Text Style Transfer / Text Formalization
- Domain: Ulasan produk e-commerce Indonesia
- Dataset: 3.000 pasang kalimat (PRDECT-ID + anotasi manual)
- Training: 19 epoch (early stopping pada epoch 14 konvergensi)
Hasil Evaluasi (Data Test 300 kalimat)
| Metrik | Skor |
|---|---|
| BLEU | 45.41 |
| ROUGE-1 | 0.7918 |
| ROUGE-2 | 0.6420 |
| ROUGE-L | 0.7837 |
| METEOR | 0.6795 |
Perbandingan dengan Baseline
| Metrik | Model Ini (T5 Fine-tuned) | Baseline (IndoNanoT5) |
|---|---|---|
| BLEU | 45.41 | 24.79 |
Cara Pakai
from transformers import T5Tokenizer, T5ForConditionalGeneration
tokenizer = T5Tokenizer.from_pretrained("SateCincau/nmt-gaul-formal")
model = T5ForConditionalGeneration.from_pretrained("SateCincau/nmt-gaul-formal")
teks_gaul = "barangnya kece bgt, recommended pokoknya!"
inputs = tokenizer(
"formalkan: " + teks_gaul,
return_tensors="pt",
max_length=128
)
output = model.generate(**inputs, max_length=128, num_beams=4)
print(tokenizer.decode(output[0], skip_special_tokens=True))
Contoh
| Input (Informal) | Output (Formal) |
|---|---|
| barangnya kece bgt, recommended! | Barang yang diterima sangat bagus. |
| seller fast respon, thx udh sabar | Penjual merespons cepat, terima kasih. |
| worth it bgt harganya | Harga sangat sepadan. |
| ongkirnya mahal tp barang cepet nyampe | Biaya pengiriman mahal namun barang cepat tiba. |
Demo Aplikasi
🌐 https://nmt-gaul-formal.streamlit.app
References
- Vaswani, A., et al. (2017). Attention Is All You Need. NeurIPS.
- Raffel, C., et al. (2020). Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer. JMLR.
- Sutoyo, R., Achmad, S., Chowanda, A., Andangsari, E. W., & Isa, S. M. (2022). PRDECT-ID: Indonesian Product Reviews Dataset for Emotions Classification Tasks. Data in Brief, 44, 108554.
- Wibowo, H. A., et al. (2020). Semi-Supervised Low-Resource Style Transfer of Indonesian Informal to Formal Language with Iterative Forward-Translation. IALP.
- Firdausillah, F., Luthfiarta, A., & Nugraha, A. T. (2025). Comparison of IndoNanoT5 and IndoGPT for Advancing Indonesian Text Formalization in Low-Resource Settings. JUTIF, 6(5), 3323–3335.
- Type
- t5
- License
- mit
- Languages
- id
- Created
- Aug 25, 2026
- Updated
- Sep 19, 2026