Live distillation demo

Black-Box Knowledge Distillation

Black-box distillation lab where a compact PyTorch student learns from teacher-labeled job postings, serves predictions through FastAPI, and reports fidelity, latency, confusion-matrix, and model-card artifacts.

Compare a compact student model against teacher-labeled job postings.

Python
PyTorch
FastAPI
React
Docker

Page expectations

What this route loads and what you should see.

Data source

Loads fidelity and cache metrics before prediction.

Expected result

Returns predicted role, probabilities, fidelity, latency, and cache metrics.

Predict with student

Compare a compact student model against teacher-labeled job postings.

Returns predicted role, probabilities, fidelity, latency, and cache metrics.

Use at least 40 characters so the model has context.

Student prediction

Returns predicted role, probabilities, fidelity, latency, and cache metrics.

Submit the form to see the prediction here.

Fidelity and cache metrics

Loads fidelity and cache metrics before prediction.

Loading fidelity and cache metrics.