securecommerce/phishing-email-detection-onnx

Phishing Email Detection (ONNX)

ONNX-converted version of dima806/phishing-email-detection for use with Transformers.js.

Model Description

DistilBERT model fine-tuned for detecting phishing emails.

• Format: ONNX (converted from PyTorch)
• Task: Binary classification (PHISHING vs LEGITIMATE)
• Accuracy: ~97%+

Usage with Transformers.js

import { pipeline } from '@xenova/transformers';

const classifier = await pipeline(
  'text-classification',
  'securecommerce/phishing-email-detection-onnx',
  { quantized: true }
);

const result = await classifier(
  'Urgent! Verify your account: http://suspicious-link.com'
);

console.log(result);
// [{ label: 'PHISHING', score: 0.98 }]

Labels

• LEGITIMATE: Safe email
• PHISHING: Phishing attempt

Credits

Original model by dima806