MCP Sentiment

🚀 Sentiment + Sarcasm Analyzer (Gradio + MCP)

This lightweight Gradio application leverages Hugging Face Transformers to perform sentiment analysis and sarcasm detection. It runs on CPU and is developed as part of the Hugging Face MCP Course. The app is fully compatible with the Hugging Face MCP server architecture.

🚀 Quick Start

This project offers a live demo for you to experience the sentiment and sarcasm analysis features right away.

👉 Launch the app on Hugging Face Spaces

✨ Features

• Sentiment Classification: Classify text as "positive" or "negative".
• Sarcasm Detection: Detect sarcasm with a probability score.
• CPU Compatibility: Run on CPU without the need for a GPU.
• Simple Interface: A straightforward and clean Gradio UI.

📦 Installation

1. Clone the repository

git clone https://github.com/YOUR_USERNAME/mcp-sentiment
cd mcp-sentiment

2. Create a virtual environment

python -m venv .venv
# Then activate:
.venv\Scripts\activate      # Windows
source .venv/bin/activate     # macOS/Linux

3. Install dependencies

pip install -r requirements.txt

Ensure gradio[mcp] is included for MCP compatibility.

4. Add Hugging Face token

Create a .env file:

HF_TOKEN=your_token_here

5. Run the app locally

python app.py

💻 Usage Examples

Basic Usage

The app returns a structured JSON response with four fields:

{
  "assessment": "positive",
  "confidence": 1.0,
  "sarcasm_detected": true,
  "sarcasm_confidence": 0.97
}

Advanced Usage

The Gradio interface provides the following controls for interaction:

📚 Documentation

Architecture Overview

• Models (CPU-only):
  • distilbert-base-uncased-finetuned-sst-2-english: For sentiment analysis.
  • helinivan/english-sarcasm-detector: For sarcasm detection.
• Frontend: Gradio UI
• Backend: Python with Hugging Face Transformers
• MCP Integration: Compatible with Hugging Face MCP (gradio[mcp])

Deploy to Hugging Face Spaces

git init
git remote add origin https://huggingface.co/spaces/YOUR_USERNAME/mcp-sentiment
git add .
git commit -m "Deploy MCP app"
git push -u origin main

Once pushed, the MCP server endpoint will be live at:

https://YOUR_USERNAME-mcp-sentiment.hf.space/gradio_api/mcp/sse

📄 License

Credits

• Hugging Face MCP Course
• Model: distilbert-base-uncased-finetuned-sst-2-english
• Model: helinivan/english-sarcasm-detector
• Gradio