README.md

metadata

title: Emotion Classifier
emoji: 🎭
colorFrom: indigo
colorTo: blue
sdk: streamlit
sdk_version: 1.51.0
app_file: app.py
pinned: false

🎭 Emotion Classifier — RoBERTa Based

A deep-learning model that classifies text into five emotion categories: anger, fear, joy, sadness, surprise.

Built with PyTorch, RoBERTa transformer, Streamlit UI and deployed on Hugging Face Spaces.

---

🧠 Model Details

Component	Description
Base model	roberta-base
Task	Single-label Emotion Classification
Input	Raw text
Output	Softmax probability distribution (5 emotions)
Framework	PyTorch + Transformers

Load Model from Hugging Face

from huggingface_hub import hf_hub_download
import torch
from BertEmotionClassifier import BertEmotionClassifier

model_path = hf_hub_download(repo_id="aadhi3/RoBert_Model", filename="model.pth")

from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("roberta-base")

model = BertEmotionClassifier(model_name="roberta-base", num_labels=5)
state_dict = torch.load(model_path, map_location="cpu")
model.load_state_dict(state_dict)
model.eval()