This is a fine-tuned IndoBERT-based model for emotion classification in Bahasa Indonesia, designed to classify everyday informal or casual text into one of five emotional categories in bahasa Indonesia. The model is particularly useful for applications involving sentiment analysis, mental health monitoring, journaling platforms, or conversational AI.
🧠 Model Overview
Base Model: indobert-base-p1 (by IndoNLU)
Fine-tuning Task: Multi-class text classification
Number of Classes: 5
Model Format: safetensors
Tokenizer: indobert-base-p1
🏷️ Emotion Classes
The model is trained to classify input text into one of the following five emotions:
Emotion
Description
Marah
Angry, Frustation
Sedih
Sadness, disappointment
Senang
Joy, excitement
Stress
Anxiety, mental pressure
Bersyukur
Gratitude, thankfulness
⚠️ Limitations
Best performance on informal Indonesian (conversational, daily tone)
May struggle with sarcasm, code-switching (mix of Indonesian-English), or domain-specific jargon
