Perth0603/Random-Forest-Model-for-PhishingDetection

Hugging Face Space - Phishing Text Classifier (Docker + FastAPI)

This Space exposes two endpoints so the Flutter app can call them reliably:

• —/predict for text/email/SMS classification via Transformers. Returns { label, score } where score is the confidence for the predicted label.
• —/predict-url for URL classification via your URL model. Returns { label, score, phishing_probability, backend, threshold } where:
• —phishing_probability is always the raw probability of phishing (0..1)
• —label is PHISH when phishing_probability >= threshold, else LEGIT
• —score is the confidence for the predicted label (for LEGIT, score = 1 - phishing_probability), which lets the app show "Safe Confidence" for legitimate URLs

Files

• —Dockerfile - builds a small FastAPI server image
• —app.py - FastAPI app that loads the model and returns normalized responses as above.
• —requirements.txt - Python dependencies.

How to deploy

1. 1.Create a new Space on Hugging Face (type: Docker).
2. 2.Upload the contents of this hf_space/ folder to the Space root (including Dockerfile).
3. 3.In Space Settings → Variables, add:
4. 4.MODEL_ID = Perth0603/phishing-email-mobilebert
5. 5.URL_REPO = Perth0603/Random-Forest-Model-for-PhishingDetection
6. 6.URLFILENAME = urlrf_model.joblib (set to your artifact filename)
7. 7.Wait for the Space to build and become green. Test:
8. 8.GET / should return { status: ok, model: ... }
9. 9.POST /predict with { "inputs": "Win an iPhone! Click here" }
10. 10.POST /predict-url with { "url": "https://example.com/login" }

Flutter app config

Set the Space URL in your env file so the app targets the Space instead of the Hosted Inference API:

{"HF_SPACE_URL":"https://<your-space>.hf.space"}

Run the app:

flutter run --dart-define-from-file=hf.env.json