# facebookresearch/mbrl-lib

Library for Model Based RL

Repository: https://github.com/facebookresearch/mbrl-lib
Canonical: https://ross.abutalabs.com/products/mbrl-lib
Language: Python
License: MIT
License Family: permissive
Topics: reinforcement-learning, model-based-reinforcement-learning
Archived: true
Last push: 2024-07-12T20:42:46+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2207, "days_push": 782, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1064, forks 173 (observed 2026-08-28T04:03:26.538351+00:00)

## What it is
MBRL-Lib is a PyTorch-based toolbox from Facebook Research for developing model-based reinforcement learning algorithms. It provides interchangeable modeling and planning components, utility functions, and reference implementations of PETS, MBPO, and PlaNet.

## Use cases
- implement a model-based RL algorithm with minimal code
- run PETS on continuous control environments like cartpole
- experiment with interchangeable dynamics models and planners
- reproduce MBPO or PlaNet research results
- prototype planning-based RL agents in PyTorch

## When to choose
- you want to develop or research model-based RL algorithms in Python
- you need ready-made implementations of PETS, MBPO, or PlaNet
- you want modular, swappable model and planning components built on PyTorch

## When to avoid
- you need model-free RL algorithms like SAC or PPO
- you need production RL deployment rather than research prototyping
- you need a library with frequent updates and active maintenance

## Facets
- artifact type: library
- maturity: maintenance
- function: reinforcement-learning, machine-learning, simulation
- domain: reinforcement-learning, machine-learning
- platform: python
- tags: model-based-rl, pytorch, pets, mbpo, planet, planning, dynamics-models, hydra, algorithms

## Member repositories
- facebookresearch/mbrl-lib (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:26.538351+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 2026-08-30T06:56:03.992069+00:00, confidence not recorded.
  - readme: https://github.com/facebookresearch/mbrl-lib (fetched 2026-08-28T04:03:26.538351+00:00, sha e6fb2cd85997)
- Data as of 2026-08-30T08:39:29.467469+00:00.