# AberHu/Knowledge-Distillation-Zoo

Pytorch implementation of various Knowledge Distillation (KD) methods.

Repository: https://github.com/AberHu/Knowledge-Distillation-Zoo
Canonical: https://ross.abutalabs.com/products/knowledge-distillation-zoo
Language: Python
License Family: other
Topics: kd-methods, knowledge-transfer, knowledge-distillation, kd, teacher-student, model-compression, distillation
Last push: 2021-11-25T06:00:37+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2756, "days_push": 1742, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1759, forks 267 (observed 2026-08-28T04:05:32.507629+00:00)

## What it is
A PyTorch reference collection implementing many classic knowledge distillation methods (logits, soft target, attention transfer, FitNet, PKT, RKD, etc.). It serves as a study and research baseline rather than a production library.

## Use cases
- implement knowledge distillation in pytorch
- compress a large model into a smaller student network
- compare different KD loss functions
- learn how teacher-student distillation works
- reproduce classic KD papers like attention transfer or FitNet
- find a starting point for model compression research

## When to choose
- you want simple, readable reference implementations of classic KD methods
- you are doing research or learning about knowledge distillation
- you need baseline KD losses to build on in PyTorch

## When to avoid
- you need a production-ready, maintained distillation framework
- you need advanced techniques like data-free, online, or self-distillation
- you require a licensed dependency in your project (no license is provided)

## Facets
- artifact type: learning-resource
- maturity: maintenance
- function: machine-learning, deep-learning, llm-training
- domain: deep-learning, machine-learning, developer-tools
- platform: python
- tags: knowledge-distillation, model-compression, pytorch, teacher-student, reference-implementation

## Member repositories
- AberHu/Knowledge-Distillation-Zoo (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:32.507629+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:27:15.558178+00:00, confidence not recorded.
  - readme: https://github.com/AberHu/Knowledge-Distillation-Zoo (fetched 2026-08-28T04:05:32.507629+00:00, sha dc7f1bf13904)
- Data as of 2026-08-30T08:39:29.467469+00:00.