# galilai-group/stable-worldmodel

A platform for reproducible world model research and evaluation

Repository: https://github.com/galilai-group/stable-worldmodel
Canonical: https://ross.abutalabs.com/products/stable-worldmodel
Homepage: https://galilai-group.github.io/stable-worldmodel/
Language: Python
License: MIT
License Family: permissive
Topics: deep-learning, pytorch, jepa, world-model, model-predictive-control
Last push: 2026-08-25T15:11:30+00:00

## Health v2 (maintenance only)
Score: 81/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 99, release rhythm 87, longevity 30
- inputs: {"age_days": 433, "days_push": 8, "days_rel": 88, "gap_med": 19, "n_releases_24m": 6}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2156, forks 263 (observed 2026-08-28T04:06:19.557699+00:00)

## What it is
A Python library providing a unified platform for reproducible world model research, covering data collection, training, and evaluation via model-predictive control across standardized environments. It ships reference baselines and planning solvers so researchers can focus on their model and objective.

## Use cases
- train and evaluate world models in standardized environments
- run model-predictive control planning baselines
- collect and store simulation datasets with Lance format
- benchmark JEPA-style world models reproducibly
- add custom environments and train world models on them
- embed planning solvers in robotics stacks

## When to choose
- you need a unified pipeline for world model data collection, training, and MPC evaluation
- you want standardized environments and reference baselines instead of re-implementing them
- you need reproducible evaluation protocols for world model research

## When to avoid
- you need a production agent framework rather than a research library
- you only need off-the-shelf pretrained world models without training
- you cannot install GPU/PyTorch dependencies

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, simulation, benchmarking, etl
- domain: machine-learning, deep-learning, reinforcement-learning, robotics
- platform: python, cross-platform
- tags: world-models, jepa, model-predictive-control, pytorch, reinforcement-learning, evaluation-benchmarks, lance-datasets, planning-solvers, research, gpu

## Member repositories
- galilai-group/stable-worldmodel (main) score 81

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:19.557699+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 2026-08-30T02:50:28.068065+00:00, confidence not recorded.
  - readme: https://github.com/galilai-group/stable-worldmodel (fetched 2026-08-28T04:06:19.557699+00:00, sha cf920411c932)
  - homepage: https://galilai-group.github.io/stable-worldmodel/ (fetched 2026-08-29T10:30:31.481922+00:00, sha c5576e7597b0)
  - registry_pypi: https://pypi.org/pypi/stable-worldmodel/json (fetched 2026-08-29T10:30:31.491550+00:00, sha 5379a316da94)
- Data as of 2026-08-30T08:39:29.467469+00:00.