app.py

from flask import Flask, request, jsonify
from transformers import BertTokenizer, BertForSequenceClassification, pipeline

# Initialize Flask app
app = Flask(__name__)

# Load pre-trained model and tokenizer
model_name = "bert-base-uncased"
tokenizer = BertTokenizer.from_pretrained(model_name)
model = BertForSequenceClassification.from_pretrained(model_name)

# Set up a pipeline
nlp_pipeline = pipeline("sentiment-analysis", model=model, tokenizer=tokenizer)

def analyze_sentiment(text):
    """
    Analyze the sentiment of the input text using the NLP pipeline.
    Returns a tuple of sentiment label and confidence score.
    """
    result = nlp_pipeline(text)
    sentiment = result[0]['label']
    confidence = result[0]['score']
    return sentiment, confidence

@app.route('/analyze', methods=['POST'])
def analyze():
    """
    API endpoint to analyze sentiment. Expects a JSON payload with 'text'.
    """
    if request.is_json:
        data = request.json
        text = data.get('text', '')
        if text:
            sentiment, confidence = analyze_sentiment(text)
            response = {
                "sentiment": sentiment,
                "confidence": confidence
            }
            return jsonify(response), 200
        else:
            return jsonify({"error": "No text provided"}), 400
    return jsonify({"error": "Invalid request format, JSON expected"}), 400

if __name__ == '__main__':
    app.run(debug=True)