fdp-flask-api / flask_api.py
dkvbabu
Clean deploy - no binary files
bf751d0
Raw History Blame Contribute Delete
1.81 kB
from flask import Flask, request, jsonify
from flask_cors import CORS
import joblib
import os
app = Flask(__name__)
CORS(app) # Allows Streamlit Space to call this API
# Load model
model = joblib.load('sentiment_model.pkl')
@app.route('/', methods=['GET'])
def home():
return jsonify({
'message': 'FDP Sentiment API is running on Hugging Face!',
'endpoints': ['/predict POST', '/batch_predict POST', '/health GET']
})
@app.route('/health', methods=['GET'])
def health():
return jsonify({'status': 'healthy', 'model': 'TF-IDF + Logistic Regression'})
@app.route('/predict', methods=['POST'])
def predict():
data = request.get_json()
text = data.get('text', '').strip()
if not text:
return jsonify({"error": "Empty text provided"}), 400
if len(text) < 3:
return jsonify({"error": "Text too short"}), 400
pred_label = model.predict([text])[0]
confidence = 0.95
if hasattr(model, 'predict_proba'):
try:
proba = model.predict_proba([text])[0]
classes = list(model.classes_)
idx = classes.index(str(pred_label))
confidence = float(proba[idx])
except Exception:
pass
return jsonify({
"sentiment": str(pred_label),
"confidence": round(confidence, 4)
})
@app.route('/batch_predict', methods=['POST'])
def batch():
data = request.get_json()
preds = model.predict(data['texts'])
probas = model.predict_proba(data['texts'])
results = [
{'text': t, 'sentiment': p, 'confidence': float(round(max(pr), 4))}
for t, p, pr in zip(data['texts'], preds, probas)
]
return jsonify({'results': results})
if __name__ == '__main__':
app.run(host='0.0.0.0', port=7860) # HF Spaces requires port 7860