galilai-group/stable-worldmodel
A platform for reproducible world model research and evaluation
Build with Backblaze B2
SDKs, agent skills, IDE extensions, and reference pipelines from Backblaze Labs. All open source.
Loading star history...
Where it stands
Rank, growth pace and release cadence relative to everything GStars tracks.
No releases in the last 90 days; the most recent was 143 days ago, with 38 contributors on GitHub.
Use Cases & Benefits
- Provides a unified platform for collecting data, training, and evaluating world models with model-predictive control across diverse standardized environments.
- Enables reproducible research with reference implementations, multiple dataset formats, and extensive environment support including factors of variation for zero-shot generalization.
- Use for developing and benchmarking world models in reinforcement learning research with standardized datasets and evaluation protocols.
- Use for training and evaluating model-predictive control policies in robotics and control tasks using a variety of simulated environments.
- Use for converting and managing large-scale datasets in different formats to optimize storage and throughput for machine learning workflows.
Should you depend on it?
License terms, project hygiene and risks, from GitHub metadata and GStars' snapshots.
License
MIT
Permissive: commercial use, modification and redistribution are allowed with attribution.
Community files
- README: present
Watch-outs
No watch-outs from GStars' checks.
General guidance from repository metadata, not legal advice. How we score
From the README
A platform for reproducible world model research and evaluation.
stable-worldmodelprovides a single, unified interface for the three stages of world model research — collecting data, training, and evaluating with model-predictive control — across a large suite of standardized environments. It ships with reference implementations of common baselines and planning solvers so research code can stay focused on the contribution that matters: the model and the objective.LeRobot dataset support is a separate opt-in extra (requires Python 3.12+):
pip install 'stable-worldmodel[lerobot]'.Datasets and checkpoints are stored under
$STABLEWM_HOME(defaults to~/.stable_worldmodel/). Override the variable to point at your preferred storage location.The library is in active development. APIs may change between minor versions.
Reference implementations are provided in
scripts/train/:lewm.pyimplements LeWM, andprejepa.pyreproduces DINO-WM. To train directly from HuggingFace object storage with no local dataset download, see the notebook.GPU utilization for LeWM trained with Push-T LanceDB dataset on a H200 GPU.
Daily snapshots
Stars gained per day, and how forks and open issues moved, from GStars' daily snapshots
Forks
289+37
Open issues
38+18
Compared with peers in Deep Learning
Fastest-growing repositories GStars tracks in this cohort for the week of Nov 1, by 7-day star growth
| Repository | Stars | +7d | Health | Last release |
|---|---|---|---|---|
| onnx/onnx | 21.6K | — | 63 | 31 days ago |
| mlc-ai/web-llm | 19.3K | — | 59 | 16 days ago |
| lukas-blecher/LaTeX-OCR | 16.6K | — | 18 | — |
| milesial/Pytorch-UNet | 11.7K | — | 18 | — |
| SharpAI/DeepCamera | 3.1K | — | 31 | 227 days ago |
| galilai-group/stable-worldmodel | 2.3K | — | 40 | 143 days ago |
| Category median | 16.6K | — | 31 | — |
Discover Repositories
Search across tracked repositories by name or description