# cbfinn/maml

Code for "Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks"

Repository: https://github.com/cbfinn/maml
Canonical: https://ross.abutalabs.com/products/maml
Language: Python
License: MIT
License Family: permissive
Last push: 2020-01-19T22:05:37+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3364, "days_push": 2418, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2727, forks 621 (observed 2026-08-28T04:07:15.932993+00:00)

## What it is
Reference implementation of Model-Agnostic Meta-Learning (MAML) from Finn et al., ICML 2017, for fast adaptation of deep networks. It includes few-shot supervised learning experiments on sinusoid regression, Omniglot classification, and MiniImagenet classification using TensorFlow.

## Use cases
- implement few-shot learning with MAML
- reproduce MAML paper experiments
- learn meta-learning algorithms
- run few-shot classification on Omniglot or MiniImagenet
- adapt a deep network to new tasks with few examples

## When to choose
- you want the canonical reference implementation of MAML for supervised few-shot learning
- you are doing research on meta-learning and need a baseline
- you want to reproduce the ICML 2017 paper results

## When to avoid
- you need MAML for reinforcement learning (use the separate maml_rl codebase)
- you need a maintained production framework or modern PyTorch support
- you need up-to-date TensorFlow compatibility

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning
- domain: machine-learning, deep-learning
- platform: python
- tags: meta-learning, few-shot-learning, tensorflow, research-code, maml, algorithms

## Member repositories
- cbfinn/maml (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:15.932993+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 2026-08-30T02:14:43.710383+00:00, confidence not recorded.
  - readme: https://github.com/cbfinn/maml (fetched 2026-08-28T04:07:15.932993+00:00, sha b25a4124885f)
- Data as of 2026-08-30T08:39:29.467469+00:00.