import csv import io import os from flask import Flask, render_template, request, jsonify import torch import torch.nn.functional as F from transformers import ( AutoTokenizer, AutoModelForSequenceClassification ) # ===================================================== # FLASK APP # ===================================================== app = Flask(__name__) # ===================================================== # DEVICE # ===================================================== device = torch.device("cpu") # ===================================================== # LOAD MODEL # ===================================================== MODEL_PATH = "./" print("Loading model...") tokenizer = AutoTokenizer.from_pretrained( MODEL_PATH ) model = AutoModelForSequenceClassification.from_pretrained( MODEL_PATH ) model.to(device) model.eval() print("Model loaded successfully!") # ===================================================== # PREDICTION FUNCTION # ===================================================== def predict_spam(text): encoding = tokenizer( text, return_tensors='pt', truncation=True, padding=True, max_length=128 ) input_ids = encoding['input_ids'].to(device) attention_mask = encoding['attention_mask'].to(device) with torch.no_grad(): outputs = model( input_ids=input_ids, attention_mask=attention_mask ) probs = F.softmax( outputs.logits, dim=1 ) spam_score = probs[0][1].item() ham_score = probs[0][0].item() pred = torch.argmax( outputs.logits, dim=1 ).item() result = "SPAM" if pred == 1 else "HAM" return { "result": result, "spam_score": round(spam_score * 100, 2), "ham_score": round(ham_score * 100, 2) } def pick_message_column(fieldnames): if not fieldnames: return None preferred_names = [ "message", "email", "text", "content", "body", "subject", "mail" ] lower_map = { name.lower().strip(): name for name in fieldnames if name } for name in preferred_names: if name in lower_map: return lower_map[name] return fieldnames[0] def analyze_csv_file(file_storage, max_rows=500): raw = file_storage.read() try: content = raw.decode("utf-8-sig") except UnicodeDecodeError: content = raw.decode("latin-1") stream = io.StringIO(content) sample = stream.read(2048) stream.seek(0) try: dialect = csv.Sniffer().sniff( sample, delimiters=",;\t|" ) except csv.Error: dialect = csv.excel reader = csv.DictReader(stream, dialect=dialect) message_column = pick_message_column(reader.fieldnames) if not message_column: return None, "CSV file must contain at least one column." rows = [] spam_count = 0 ham_count = 0 for index, row in enumerate(reader, start=1): if index > max_rows: break message = (row.get(message_column) or "").strip() if not message: continue prediction = predict_spam(message) if prediction["result"] == "SPAM": spam_count += 1 else: ham_count += 1 rows.append({ "row": index, "message": message, "preview": message[:140], **prediction }) total = len(rows) return { "message_column": message_column, "total": total, "spam": spam_count, "ham": ham_count, "spam_rate": round((spam_count / total) * 100, 2) if total else 0, "ham_rate": round((ham_count / total) * 100, 2) if total else 0, "rows": rows }, None # ===================================================== # ROUTES # ===================================================== @app.route('/') def home(): return render_template( 'index.html' ) @app.route('/predict', methods=['POST']) def predict(): data = request.get_json(silent=True) or {} text = data.get( 'message', '' ).strip() if not text: return jsonify({ "error": "Message text is required." }), 400 result = predict_spam(text) return jsonify(result) @app.route('/predict_csv', methods=['POST']) def predict_csv(): csv_file = request.files.get("file") if not csv_file or not csv_file.filename: return jsonify({ "error": "CSV file is required." }), 400 if not csv_file.filename.lower().endswith(".csv"): return jsonify({ "error": "Please upload a .csv file." }), 400 result, error = analyze_csv_file(csv_file) if error: return jsonify({ "error": error }), 400 return jsonify(result) # ===================================================== # RUN APP # ===================================================== if __name__ == '__main__': app.run( host='0.0.0.0', port=int(os.environ.get("PORT", 7860)), debug=False )