{
    "info": {
        "author": "Illarion Khlestov",
        "author_email": "khlyestovillarion@gmail.com",
        "bugtrack_url": null,
        "classifiers": [
            "Intended Audience :: Developers",
            "Intended Audience :: Science/Research",
            "License :: OSI Approved :: Apache Software License",
            "Programming Language :: Python :: 3.4",
            "Programming Language :: Python :: 3.5",
            "Programming Language :: Python :: 3.6",
            "Programming Language :: Python :: 3.7",
            "Topic :: Scientific/Engineering :: Information Analysis",
            "Topic :: Software Development :: Libraries"
        ],
        "description": "===================================================\nMonitoring utility for machine learning experiments\n===================================================\n\n.. image:: https://travis-ci.com/ikhlestov/rarog.svg?branch=master\n   :target: https://travis-ci.com/ikhlestov/rarog/\n   :alt: Travis status for master branch\n\n.. image:: https://codecov.io/gh/ikhlestov/rarog/branch/master/graph/badge.svg\n   :target: https://codecov.io/gh/ikhlestov/rarog/\n   :alt: codecov.io status for master branch\n\n.. image:: https://img.shields.io/pypi/pyversions/rarog.svg\n    :target: https://pypi.org/project/rarog\n\n\nRarog is a monitoring utility for machine learning experiments. You may use it as a\nlight-weight alternative for `TensorBoard <https://github.com/tensorflow/tensorboard>`_\nor `Visdom <https://github.com/facebookresearch/visdom>`_. Rarog stores all records in\n`ClickHouse`_ database using \n`ClickHouse Python Driver <https://github.com/mymarilyn/clickhouse-driver>`__.\n\nFeatures\n========\n\n- log common python data types(bool, int, float, string, iterables)\n- log 1d numpy arrays\n- distributed experiments monitoring\n\nSetup\n=====\n\nInstall via `pip`_:\n\n.. code-block:: bash\n\n    pip install rarog\n\nStart `ClickHouse`_ database if required:\n\n.. code-block:: bash\n\n    docker run -d --name clickhouse --ulimit nofile=262144:262144 -p 9000:9000 yandex/clickhouse-server\n\n**Important note:** the example above is just the easiest way. For production, you should\nsetup database backups or replicated.\n\nRarog supports Python 3.4 and newer.\n\nUsage\n===============\n\n.. code:: python3\n\n    import random\n    \n    from rarog import Tracker\n    \n    tracker = Tracker(name='experiment_name')\n    \n    # trace values one by one\n    for step in range(10):\n        tracker.trace(\n            name='int_value',\n            value=random.randint(10, 20),\n            step=step)\n        tracker.trace(\n            name='float_value',\n            value=random.random(),\n            step=step)\n        # provide experiment phase as a string\n        tracker.trace(\n            name='list_value',\n            value=[random.random(), random.random()],\n            step=step,\n            phase='val')\n        \n    # trace values by dict\n    for step in range(10, 20):\n        tracker.multy_trace({\n            'int_value': random.randint(10, 20),\n            'float_value': random.random()\n        }, step=step)\n    \n    # get names of traced metrics\n    tracker.metrics\n    # Out: ['time', 'step', 'phase', 'int_value', 'float_value', 'list_value']\n\n\nIf you are going to record more than 100 entries per second,\nit's better to use ``sync_step`` or ``sync_seconds`` arguments.\nThus writing to the database will be done with some period, which is much faster.\n\n.. code:: python3\n\n    # `exist_ok` flag allow to use the same name for experiment\n    step_tracker = Tracker(name='experiment_name', sync_step=1000, exist_ok=True)\n    \n    for step in range(20, 10**4):\n        step_tracker.trace(name='bool_value', value=bool(random.randint(0, 1)), step=step)\n        step_tracker.multy_trace({\n            'int_value': random.randint(10, 20),\n            'float_value': random.random()\n        }, step=step)\n    \n    # tracker should be manually synchronized after last entry\n    step_tracker.sync_accumulated_values()\n\nExperiments can be handled via manager\n\n.. code:: python3\n\n    from rarog import Manager\n    \n    manager = Manager()\n    manager.list_experiments()\n    # Out: ['experiment_name']\n    \n    manager.remove_experiment('experiment_name')\n\n\n\nRetrieving your data\n====================\n\nTODO (manually and with visualization)\n\n\n.. _ClickHouse: https://clickhouse.yandex\n.. _pip: https://pip.pypa.io/en/stable/quickstart/\n\n\nTODO\n=====\n\n- Pytorch tensors support\n- Support 2d arrays\n- Tensorflow data types support\n- Split Aggregation View for summarization and underlying tables\n- Store experiments metadata(config, author, etc.)\n- Autodocs",
        "description_content_type": "",
        "docs_url": null,
        "download_url": "",
        "downloads": {
            "last_day": -1,
            "last_month": -1,
            "last_week": -1
        },
        "home_page": "https://github.com/ikhlestov/rarog",
        "keywords": "",
        "license": "Apache 2",
        "maintainer": "Illarion Khlestov",
        "maintainer_email": "khlyestovillarion@gmail.com",
        "name": "rarog",
        "package_url": "https://pypi.org/project/rarog/",
        "platform": "",
        "project_url": "https://pypi.org/project/rarog/",
        "project_urls": {
            "CI: Travis": "https://travis-ci.com/ikhlestov/rarog/",
            "Coverage: codecov": "https://codecov.io/gh/ikhlestov/rarog/",
            "GitHub: repo": "https://github.com/ikhlestov/rarog",
            "Homepage": "https://github.com/ikhlestov/rarog"
        },
        "release_url": "https://pypi.org/project/rarog/0.1.dev1/",
        "requires_dist": null,
        "requires_python": ">=3.4.0",
        "summary": "Monitoring utility for machine learning experiments",
        "version": "0.1.dev1"
    },
    "last_serial": 5757442,
    "releases": {
        "0.1.dev0": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "b5da099264da21dc1ad17942282e4464",
                    "sha256": "e2b2625f407ac6a15efbd81b13843a4ef67c24e80b96b915335201a645ed3446"
                },
                "downloads": -1,
                "filename": "rarog-0.1.dev0.tar.gz",
                "has_sig": false,
                "md5_digest": "b5da099264da21dc1ad17942282e4464",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": ">=3.4.0",
                "size": 5643,
                "upload_time": "2019-08-29T22:34:43",
                "url": "https://files.pythonhosted.org/packages/14/1e/f786b6aca6aa4ac97d7922a9ca49bb3cbf7622d2383053d0a866b6cebb2b/rarog-0.1.dev0.tar.gz"
            }
        ],
        "0.1.dev1": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "066c3b7b7ce00613c35b4900096cb46e",
                    "sha256": "71351edd1bd9391172fa86b36c64fab58f67df8e3d1e73940a449de6a24ea473"
                },
                "downloads": -1,
                "filename": "rarog-0.1.dev1.tar.gz",
                "has_sig": false,
                "md5_digest": "066c3b7b7ce00613c35b4900096cb46e",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": ">=3.4.0",
                "size": 5909,
                "upload_time": "2019-08-29T23:10:47",
                "url": "https://files.pythonhosted.org/packages/05/b3/0e9e69a8aa471c236c4dc4f7db61d400cd218959972abb1f0cd0437df240/rarog-0.1.dev1.tar.gz"
            }
        ]
    },
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "md5": "066c3b7b7ce00613c35b4900096cb46e",
                "sha256": "71351edd1bd9391172fa86b36c64fab58f67df8e3d1e73940a449de6a24ea473"
            },
            "downloads": -1,
            "filename": "rarog-0.1.dev1.tar.gz",
            "has_sig": false,
            "md5_digest": "066c3b7b7ce00613c35b4900096cb46e",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": ">=3.4.0",
            "size": 5909,
            "upload_time": "2019-08-29T23:10:47",
            "url": "https://files.pythonhosted.org/packages/05/b3/0e9e69a8aa471c236c4dc4f7db61d400cd218959972abb1f0cd0437df240/rarog-0.1.dev1.tar.gz"
        }
    ]
}