xanderabim/fraudfoxai-phishing

FraudFoxAI Phishing Detection Model

Fine-tuned DistilBERT model for detecting phishing and fraudulent emails. Trained on 214,000 emails from CEAS_08 and Phishing_Email datasets.

Model Details

• Base Model: distilbert-base-uncased
• Training Data: 214K emails (CEAS_08 + Phishing_Email)
• Classes:
  • LABEL_0: Legitimate Email
  • LABEL_1: Phishing/Fraud Email

Training Configuration

• Epochs: 3
• Batch Size: 16
• Learning Rate: 2e-5
• Max Length: 512 tokens
• Framework: PyTorch + Transformers

Usage

from transformers import AutoTokenizer, AutoModelForSequenceClassification

tokenizer = AutoTokenizer.from_pretrained("xanderabim/fraudfoxai-phishing")
model = AutoModelForSequenceClassification.from_pretrained("xanderabim/fraudfoxai-phishing")

inputs = tokenizer("URGENT: Verify your account now!", return_tensors="pt", truncation=True)
outputs = model(**inputs)

Production API

Deployed at: https://fraudfoxai.xanderabim.workers.dev

curl -X POST https://fraudfoxai.xanderabim.workers.dev/api/analyze \
  -H "Content-Type: application/json" \
  -d '{"subject": "Account verification", "body": "Click here now", "from": "unknown@site.com"}'

Limitations

• English language only
• Max 512 tokens per input
• May flag aggressive marketing emails

License

MIT

Author

@xanderabim