app.py

from flask import Flask, request, jsonify
from transformers import DistilBertTokenizer, TFDistilBertForSequenceClassification
import tensorflow as tf
import re

# Initialize Flask app
app = Flask(__name__)

@app.route('/')
def health_check():
    return jsonify({"status": "healthy"}), 200

# --- Load Model and Tokenizer ---
# Load the fine-tuned model and tokenizer from the saved directory
model_path = "./saved_model"
tokenizer = DistilBertTokenizer.from_pretrained(model_path)
model = TFDistilBertForSequenceClassification.from_pretrained(model_path)

print("Model and Tokenizer loaded successfully!")

# --- Helper Function for Text Cleaning ---
def clean_text(text):
    text = text.lower()
    text = re.sub(r'@[a-zA-Z0-9_]+', '', text)
    text = re.sub(r'https?://[A-Za-z0-9./]+', '', text)
    text = re.sub(r'[^a-zA-Z\s]', '', text)
    text = re.sub(r'\s+', ' ', text).strip()
    return text

# --- Define the Prediction Endpoint ---
@app.route('/predict', methods=['POST'])
def predict():
    try:
        # Get text from the request's JSON body
        data = request.get_json()
        text = data['text']

        if not text:
            return jsonify({"error": "Text field is required"}), 400

        # Clean the input text
        cleaned_text = clean_text(text)

        # Tokenize the text
        inputs = tokenizer(cleaned_text, return_tensors="tf", truncation=True, padding=True, max_length=128)

        # Make a prediction
        outputs = model(inputs)
        logits = outputs.logits
        probabilities = tf.nn.softmax(logits, axis=-1)[0].numpy()
        prediction = tf.argmax(logits, axis=-1).numpy()[0]

        # Map prediction index to label
        labels = ['Non-Toxic', 'Toxic (Hate Speech/Offensive)']
        result_label = labels[prediction]
        confidence = float(probabilities[prediction])

        # Return the result
        return jsonify({
            "text": text,
            "prediction": result_label,
            "confidence": f"{confidence:.4f}"
        })

    except Exception as e:
        return jsonify({"error": str(e)}), 500

# Run the app
if __name__ == '__main__':
    # Use 0.0.0.0 to make it accessible from other Docker containers
    app.run(host='0.0.0.0', port=5000)