starforce


Namestarforce JSON
Version 1.1.6 PyPI version JSON
download
home_pagehttps://github.com/
SummaryNone
upload_time2025-07-10 09:15:21
maintainerNone
docs_urlNone
authorLucas Jin
requires_python>=3.10
licenseNone
keywords
VCS
bugtrack_url
requirements No requirements were recorded.
Travis-CI No Travis.
coveralls test coverage No coveralls.
            # StarForce

Dual system training framework for robotics.

The overall structure borrowed from GR00T, with these modifications:

- Simplifer VLM model, introduced more advanced VLM and larger slow system;
- Connecting fast system without cross-attention, using text encoder instead;



## Slow thought system

will goes `starforce/model/backbone` contains various VLMs. Provides a unified interface connect with fast system (action expert)



## Fast action system

Currently support:

- DiT: diffusion transformer
- QwenFlow: flowmatching based action expert



## Env install

```
pip install -e .[base]
```




## Training

training scripts goes to `scripts/xxx.sh`

training fast system:

```
sh scripts/v0/sl_0.sh

```

training slow thinking system:







            

Raw data

            {
    "_id": null,
    "home_page": "https://github.com/",
    "name": "starforce",
    "maintainer": null,
    "docs_url": null,
    "requires_python": ">=3.10",
    "maintainer_email": null,
    "keywords": null,
    "author": "Lucas Jin",
    "author_email": "\"GEAR@NVIDIA-RESEARCH\" <*@nvidia.com>",
    "download_url": "https://files.pythonhosted.org/packages/ab/9c/b814dfb58e8e26ac57cf38daf0d0666fd81101a03c06e6dd8c1813103404/starforce-1.1.6.tar.gz",
    "platform": "any",
    "description": "# StarForce\r\n\r\nDual system training framework for robotics.\r\n\r\nThe overall structure borrowed from GR00T, with these modifications:\r\n\r\n- Simplifer VLM model, introduced more advanced VLM and larger slow system;\r\n- Connecting fast system without cross-attention, using text encoder instead;\r\n\r\n\r\n\r\n## Slow thought system\r\n\r\nwill goes `starforce/model/backbone` contains various VLMs. Provides a unified interface connect with fast system (action expert)\r\n\r\n\r\n\r\n## Fast action system\r\n\r\nCurrently support:\r\n\r\n- DiT: diffusion transformer\r\n- QwenFlow: flowmatching based action expert\r\n\r\n\r\n\r\n## Env install\r\n\r\n```\r\npip install -e .[base]\r\n```\r\n\r\n\r\n\r\n\r\n## Training\r\n\r\ntraining scripts goes to `scripts/xxx.sh`\r\n\r\ntraining fast system:\r\n\r\n```\r\nsh scripts/v0/sl_0.sh\r\n\r\n```\r\n\r\ntraining slow thinking system:\r\n\r\n\r\n\r\n\r\n\r\n\r\n",
    "bugtrack_url": null,
    "license": null,
    "summary": null,
    "version": "1.1.6",
    "project_urls": {
        "Homepage": "https://github.com/"
    },
    "split_keywords": [],
    "urls": [
        {
            "comment_text": null,
            "digests": {
                "blake2b_256": "ab9cb814dfb58e8e26ac57cf38daf0d0666fd81101a03c06e6dd8c1813103404",
                "md5": "adccdd45602b7c5945499d87c5cbd099",
                "sha256": "4229cb07389445f438fe0c3c1ef1cf3577ad9287a44f0cd541e87542569f5926"
            },
            "downloads": -1,
            "filename": "starforce-1.1.6.tar.gz",
            "has_sig": false,
            "md5_digest": "adccdd45602b7c5945499d87c5cbd099",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": ">=3.10",
            "size": 1850267,
            "upload_time": "2025-07-10T09:15:21",
            "upload_time_iso_8601": "2025-07-10T09:15:21.782437Z",
            "url": "https://files.pythonhosted.org/packages/ab/9c/b814dfb58e8e26ac57cf38daf0d0666fd81101a03c06e6dd8c1813103404/starforce-1.1.6.tar.gz",
            "yanked": false,
            "yanked_reason": null
        }
    ],
    "upload_time": "2025-07-10 09:15:21",
    "github": false,
    "gitlab": false,
    "bitbucket": false,
    "codeberg": false,
    "lcname": "starforce"
}
        
Elapsed time: 0.66659s