{
    "info": {
        "author": "Illia Polosukhin, NEAR Inc",
        "author_email": "illia@near.ai",
        "bugtrack_url": null,
        "classifiers": [],
        "description": "# TorchFold\n\nBlog post: http://near.ai/articles/2017-09-06-PyTorch-Dynamic-Batching/\n\nAnalogous to [TensorFlow Fold](https://github.com/tensorflow/fold), implements dynamic batching with super simple interface.\nReplace every direct call in your computation to nn module with `f.add('function name', arguments)`.\nIt will construct an optimized version of computation and on `f.apply` will dynamically batch and execute the computation on given nn module.\n\n## Installation\nWe recommend using pip package manager:\n```\npip install torchfold\n```\n\n## Example\n\n```\n    f = torchfold.Fold()\n\n    def dfs(node):\n        if is_leaf(node):\n            return f.add('leaf', node)\n        else:\n            prev = f.add('init')\n            for child in children(node):\n                prev = f.add('child', prev, child)\n            return prev\n\n    class Model(nn.Module):\n        def __init__(self, ...):\n            ...\n\n        def leaf(self, leaf):\n            ...\n\n        def child(self, prev, child):\n            ...\n\n    res = dfs(my_tree)\n    model = Model(...)\n    f.apply(model, [[res]])\n```\n\n\n",
        "description_content_type": "text/markdown",
        "docs_url": null,
        "download_url": "",
        "downloads": {
            "last_day": -1,
            "last_month": -1,
            "last_week": -1
        },
        "home_page": "",
        "keywords": "",
        "license": "Apache License, Version 2.0",
        "maintainer": "",
        "maintainer_email": "",
        "name": "torchfold",
        "package_url": "https://pypi.org/project/torchfold/",
        "platform": "",
        "project_url": "https://pypi.org/project/torchfold/",
        "project_urls": {
            "Blog Post": "http://near.ai/articles/2017-09-06-PyTorch-Dynamic-Batching/",
            "Source": "https://github.com/nearai/torchfold"
        },
        "release_url": "https://pypi.org/project/torchfold/0.1.0/",
        "requires_dist": null,
        "requires_python": "",
        "summary": "Dynamic Batching with PyTorch",
        "version": "0.1.0"
    },
    "last_serial": 3940981,
    "releases": {
        "0.1.0": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "7777910c57b940a96f0c591a90c1ca71",
                    "sha256": "53721bbf32b61119f596b23d44061503497d5eb6fa67a3c1c82ba008842b3c26"
                },
                "downloads": -1,
                "filename": "torchfold-0.1.0-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "7777910c57b940a96f0c591a90c1ca71",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 5352,
                "upload_time": "2018-06-07T20:34:11",
                "url": "https://files.pythonhosted.org/packages/c0/fb/2ff01af27b6fbe147a16be0e3f78007a3d03224d89ebcfa2441d838398bc/torchfold-0.1.0-py3-none-any.whl"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "3ec270cac4e5f7b5219a5495d07b9a83",
                    "sha256": "73605bdbbaa627735bb28c3f90d654896de248fa1a988e730182b68ddf4660a6"
                },
                "downloads": -1,
                "filename": "torchfold-0.1.0.tar.gz",
                "has_sig": false,
                "md5_digest": "3ec270cac4e5f7b5219a5495d07b9a83",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 4703,
                "upload_time": "2018-06-07T20:34:12",
                "url": "https://files.pythonhosted.org/packages/12/80/4c88fea850af25cb66e5d07eff3b38411eaa09f94c5a3c4370a7316a9234/torchfold-0.1.0.tar.gz"
            }
        ]
    },
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "md5": "7777910c57b940a96f0c591a90c1ca71",
                "sha256": "53721bbf32b61119f596b23d44061503497d5eb6fa67a3c1c82ba008842b3c26"
            },
            "downloads": -1,
            "filename": "torchfold-0.1.0-py3-none-any.whl",
            "has_sig": false,
            "md5_digest": "7777910c57b940a96f0c591a90c1ca71",
            "packagetype": "bdist_wheel",
            "python_version": "py3",
            "requires_python": null,
            "size": 5352,
            "upload_time": "2018-06-07T20:34:11",
            "url": "https://files.pythonhosted.org/packages/c0/fb/2ff01af27b6fbe147a16be0e3f78007a3d03224d89ebcfa2441d838398bc/torchfold-0.1.0-py3-none-any.whl"
        },
        {
            "comment_text": "",
            "digests": {
                "md5": "3ec270cac4e5f7b5219a5495d07b9a83",
                "sha256": "73605bdbbaa627735bb28c3f90d654896de248fa1a988e730182b68ddf4660a6"
            },
            "downloads": -1,
            "filename": "torchfold-0.1.0.tar.gz",
            "has_sig": false,
            "md5_digest": "3ec270cac4e5f7b5219a5495d07b9a83",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": null,
            "size": 4703,
            "upload_time": "2018-06-07T20:34:12",
            "url": "https://files.pythonhosted.org/packages/12/80/4c88fea850af25cb66e5d07eff3b38411eaa09f94c5a3c4370a7316a9234/torchfold-0.1.0.tar.gz"
        }
    ]
}