indobertweet-metadata-s2

Fine-tuned dari indolem/indobertweet-base-uncased, hyperparameter dicari dengan Optuna.

Problem type: multi_label_classification
Labels: insults, identity_attack, threat_incitement_to_violence, profanity_obscenity, sexually_explicit

Penggunaan

from transformers import AutoTokenizer, AutoModelForSequenceClassification

tokenizer = AutoTokenizer.from_pretrained("evelnap/indobertweet-metadata-s2")
model = AutoModelForSequenceClassification.from_pretrained("evelnap/indobertweet-metadata-s2")

Catatan: teks input sebaiknya diproses dengan preprocessing yang sama seperti saat training (lowercase, normalisasi emoji, mention, URL, dst.) agar hasil prediksi konsisten.

Downloads last month: 38

Safetensors

Model size

0.1B params

Tensor type

F32

Model tree for evelnap/indobertweet-metadata-s2

Base model

indolem/indobertweet-base-uncased

Finetuned

(65)

this model