DeepLearning Sentiment Analysis

A Streamlit web application for performing real-time sentiment analysis using a Hugging Face Transformer model powered by PyTorch. The app classifies text as positive, negative, or neutral and displays confidence scores for each class.

Features

• Real-time sentiment prediction from user input
• Transformer-based NLP pipeline via transformers and torch
• Simple Streamlit UI with probability visualization
• Lightweight deployment that runs locally or in the cloud

Project Structure

• app.py – Streamlit application entry point that loads the model and renders the UI.
• requirements.txt – Python dependencies needed to run the app.

Getting Started

Prerequisites

• Python 3.10+ recommended
• Virtual environment tool such as venv or conda

Installation

1. Clone the repository and enter the project directory.
2. Create and activate a virtual environment:

   python -m venv .venv
   source .venv/bin/activate

3. Install dependencies:

   pip install -r requirements.txt

Running the App

Launch the Streamlit server from the project root:

streamlit run app.py

The app will provide a local URL (typically http://localhost:8501) where you can enter text and view predicted sentiment labels with confidence scores.

Deployment Notes

• Model weights are downloaded at runtime; ensure outbound network access on first launch.
• For containerized deployments, cache the model by running the app once during the image build or specify a model path via environment configuration if you extend the code.

Contributing

Issues and pull requests are welcome. Please ensure changes include appropriate tests or manual verification steps when applicable.