🧠 Emotion Classifier & Supportive Chatbot API (Flask + Together AI)

This is a Flask-based backend API that leverages Together AI and LLaMA 3 (8B Chat) model to:

1. Classify user emotions from journal entries or personal messages.
2. Respond empathetically to user "curhatan" (emotional expressions) in Bahasa Indonesia.

---

📌 Endpoints

POST /predict – Emotion Classification

Classifies a given text into one of six emotions: joy, anger, sadness, love, fear, or surprise.

✅ Request

{
  "content": "Isi curhatan atau pesan pengguna di sini"
}

🔁 Response

{
  "text": "Isi curhatan atau pesan pengguna di sini",
  "label": "joy",
  "score": 30,
  "raw_model_output": "1"
}

POST /curhat – Empathetic Response

Classifies a given text into one of six emotions: joy, anger, sadness, love, fear, or surprise.

✅ Request

{
  "message": "Aku merasa lelah dan tidak tahu harus bagaimana lagi..."
}

🔁 Response

{
  "reply": "Saya mengerti perasaanmu. Kadang kita memang butuh waktu untuk diri sendiri. Jangan menyerah, kamu tidak sendirian."
}

---

📄 Requirements

To run this project, you'll need the following Python packages:

Package	Version	Description
Flask	3.1.1	Web framework to build the API server
python-dotenv	1.1.1	Load environment variables from .env
requests	2.32.4	Make HTTP requests to Together API
gunicorn	latest	Production WSGI server for deployment

Install all dependencies using:

pip install -r requirements.txt