{
    "info": {
        "author": "Bayu Aldi Yansyah",
        "author_email": "bayualdiyansyah@gmail.com",
        "bugtrack_url": null,
        "classifiers": [
            "Development Status :: 2 - Pre-Alpha",
            "Intended Audience :: Developers",
            "Intended Audience :: Science/Research",
            "License :: OSI Approved :: BSD License",
            "Programming Language :: C++",
            "Programming Language :: Cython",
            "Programming Language :: Python :: 2.6",
            "Programming Language :: Python :: 2.7",
            "Programming Language :: Python :: 3",
            "Programming Language :: Python :: 3.2",
            "Programming Language :: Python :: 3.3",
            "Programming Language :: Python :: 3.4",
            "Programming Language :: Python :: 3.5",
            "Topic :: Scientific/Engineering :: Artificial Intelligence"
        ],
        "description": "fasttext |Build Status| |PyPI version|\n======================================\n\nfasttext is a Python interface for `Facebook\nfastText <https://github.com/facebookresearch/fastText>`__.\n\nRequirements\n------------\n\nfasttext support Python 2.6 or newer. It requires\n`Cython <https://pypi.python.org/pypi/Cython/>`__ in order to build the\nC++ extension.\n\nInstallation\n------------\n\n.. code:: shell\n\n    pip install fasttext\n\nExample usage\n-------------\n\nThis package has two main use cases: word representation learning and\ntext classification.\n\nThese were described in the two papers\n`1 <#enriching-word-vectors-with-subword-information>`__ and\n`2 <#bag-of-tricks-for-efficient-text-classification>`__.\n\nWord representation learning\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\nIn order to learn word vectors, as described in\n`1 <#enriching-word-vectors-with-subword-information>`__, we can use\n``fasttext.skipgram`` and ``fasttext.cbow`` function like the following:\n\n.. code:: python\n\n    import fasttext\n\n    # Skipgram model\n    model = fasttext.skipgram('data.txt', 'model')\n    print model.words # list of words in dictionary\n\n    # CBOW model\n    model = fasttext.cbow('data.txt', 'model')\n    print model.words # list of words in dictionary\n\nwhere ``data.txt`` is a training file containing ``utf-8`` encoded text.\nBy default the word vectors will take into account character n-grams\nfrom 3 to 6 characters.\n\nAt the end of optimization the program will save two files:\n``model.bin`` and ``model.vec``.\n\n``model.vec`` is a text file containing the word vectors, one per line.\n``model.bin`` is a binary file containing the parameters of the model\nalong with the dictionary and all hyper parameters.\n\nThe binary file can be used later to compute word vectors or to restart\nthe optimization.\n\nThe following ``fasttext(1)`` command is equivalent\n\n.. code:: shell\n\n    # Skipgram model\n    ./fasttext skipgram -input data.txt -output model\n\n    # CBOW model\n    ./fasttext cbow -input data.txt -output model\n\nObtaining word vectors for out-of-vocabulary words\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\nThe previously trained model can be used to compute word vectors for\nout-of-vocabulary words.\n\n.. code:: python\n\n    print model['king'] # get the vector of the word 'king'\n\nthe following ``fasttext(1)`` command is equivalent:\n\n.. code:: shell\n\n    echo \"king\" | ./fasttext print-vectors model.bin\n\nThis will output the vector of word ``king`` to the standard output.\n\nLoad pre-trained model\n~~~~~~~~~~~~~~~~~~~~~~\n\nWe can use ``fasttext.load_model`` to load pre-trained model:\n\n.. code:: python\n\n    model = fasttext.load_model('model.bin')\n    print model.words # list of words in dictionary\n    print model['king'] # get the vector of the word 'king'\n\nText classification\n~~~~~~~~~~~~~~~~~~~\n\nThis package can also be used to train supervised text classifiers and\nload pre-trained classifier from fastText.\n\nIn order to train a text classifier using the method described in\n`2 <#bag-of-tricks-for-efficient-text-classification>`__, we can use the\nfollowing function:\n\n.. code:: python\n\n    classifier = fasttext.supervised('data.train.txt', 'model')\n\nequivalent as ``fasttext(1)`` command:\n\n.. code:: shell\n\n    ./fasttext supervised -input data.train.txt -output model\n\nwhere ``data.train.txt`` is a text file containing a training sentence\nper line along with the labels. By default, we assume that labels are\nwords that are prefixed by the string ``__label__``.\n\nWe can specify the label prefix with the ``label_prefix`` param:\n\n.. code:: python\n\n    classifier = fasttext.supervised('data.train.txt', 'model', label_prefix='__label__')\n\nequivalent as ``fasttext(1)`` command:\n\n.. code:: shell\n\n    ./fasttext supervised -input data.train.txt -output model -label '__label__'\n\nThis will output two files: ``model.bin`` and ``model.vec``.\n\nOnce the model was trained, we can evaluate it by computing the\nprecision at 1 (P@1) and the recall on a test set using\n``classifier.test`` function:\n\n.. code:: python\n\n    result = classifier.test('test.txt')\n    print 'P@1:', result.precision\n    print 'R@1:', result.recall\n    print 'Number of examples:', result.nexamples\n\nThis will print the same output to stdout as:\n\n.. code:: shell\n\n    ./fasttext test model.bin test.txt\n\nIn order to obtain the most likely label for a list of text, we can use\n``classifer.predict`` method:\n\n.. code:: python\n\n    texts = ['example very long text 1', 'example very longtext 2']\n    labels = classifier.predict(texts)\n    print labels\n\n    # Or with the probability\n    labels = classifier.predict_proba(texts)\n    print labels\n\nWe can specify ``k`` value to get the k-best labels from classifier:\n\n.. code:: python\n\n    labels = classifier.predict(texts, k=3)\n    print labels\n\n    # Or with the probability\n    labels = classifier.predict_proba(texts, k=3)\n    print labels\n\nThis interface is equivalent as ``fasttext(1)`` predict command. The\nsame model with the same input set will have the same prediction.\n\nAPI documentation\n-----------------\n\nSkipgram model\n~~~~~~~~~~~~~~\n\nTrain & load skipgram model\n\n.. code:: python\n\n    model = fasttext.skipgram(params)\n\nList of available ``params`` and their default value:\n\n::\n\n    input_file     training file path (required)\n    output         output file path (required)\n    lr             learning rate [0.05]\n    lr_update_rate change the rate of updates for the learning rate [100]\n    dim            size of word vectors [100]\n    ws             size of the context window [5]\n    epoch          number of epochs [5]\n    min_count      minimal number of word occurences [5]\n    neg            number of negatives sampled [5]\n    word_ngrams    max length of word ngram [1]\n    loss           loss function {ns, hs, softmax} [ns]\n    bucket         number of buckets [2000000]\n    minn           min length of char ngram [3]\n    maxn           max length of char ngram [6]\n    thread         number of threads [12]\n    t              sampling threshold [0.0001]\n    silent         disable the log output from the C++ extension [1]\n    encoding       specify input_file encoding [utf-8]\n\nExample usage:\n\n.. code:: python\n\n    model = fasttext.skipgram('train.txt', 'model', lr=0.1, dim=300)\n\nCBOW model\n~~~~~~~~~~\n\nTrain & load CBOW model\n\n.. code:: python\n\n    model = fasttext.cbow(params)\n\nList of available ``params`` and their default value:\n\n::\n\n    input_file     training file path (required)\n    output         output file path (required)\n    lr             learning rate [0.05]\n    lr_update_rate change the rate of updates for the learning rate [100]\n    dim            size of word vectors [100]\n    ws             size of the context window [5]\n    epoch          number of epochs [5]\n    min_count      minimal number of word occurences [5]\n    neg            number of negatives sampled [5]\n    word_ngrams    max length of word ngram [1]\n    loss           loss function {ns, hs, softmax} [ns]\n    bucket         number of buckets [2000000]\n    minn           min length of char ngram [3]\n    maxn           max length of char ngram [6]\n    thread         number of threads [12]\n    t              sampling threshold [0.0001]\n    silent         disable the log output from the C++ extension [1]\n    encoding       specify input_file encoding [utf-8]\n\nExample usage:\n\n.. code:: python\n\n    model = fasttext.cbow('train.txt', 'model', lr=0.1, dim=300)\n\nLoad pre-trained model\n~~~~~~~~~~~~~~~~~~~~~~\n\nFile ``.bin`` that previously trained or generated by fastText can be\nloaded using this function\n\n.. code:: python\n\n    model = fasttext.load_model('model.bin', encoding='utf-8')\n\nAttributes and methods for the model\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\nSkipgram and CBOW model have the following atributes & methods\n\n.. code:: python\n\n    model.model_name       # Model name\n    model.words            # List of words in the dictionary\n    model.dim              # Size of word vector\n    model.ws               # Size of context window\n    model.epoch            # Number of epochs\n    model.min_count        # Minimal number of word occurences\n    model.neg              # Number of negative sampled\n    model.word_ngrams      # Max length of word ngram\n    model.loss_name        # Loss function name\n    model.bucket           # Number of buckets\n    model.minn             # Min length of char ngram\n    model.maxn             # Max length of char ngram\n    model.lr_update_rate   # Rate of updates for the learning rate\n    model.t                # Value of sampling threshold\n    model.encoding         # Encoding of the model\n    model[word]            # Get the vector of specified word\n\nSupervised model\n~~~~~~~~~~~~~~~~\n\nTrain & load the classifier\n\n.. code:: python\n\n    classifier = fasttext.supervised(params)\n\nList of available ``params`` and their default value:\n\n::\n\n    input_file             training file path (required)\n    output                 output file path (required)\n    label_prefix           label prefix ['__label__']\n    lr                     learning rate [0.1]\n    lr_update_rate         change the rate of updates for the learning rate [100]\n    dim                    size of word vectors [100]\n    ws                     size of the context window [5]\n    epoch                  number of epochs [5]\n    min_count              minimal number of word occurences [1]\n    neg                    number of negatives sampled [5]\n    word_ngrams            max length of word ngram [1]\n    loss                   loss function {ns, hs, softmax} [softmax]\n    bucket                 number of buckets [0]\n    minn                   min length of char ngram [0]\n    maxn                   max length of char ngram [0]\n    thread                 number of threads [12]\n    t                      sampling threshold [0.0001]\n    silent                 disable the log output from the C++ extension [1]\n    encoding               specify input_file encoding [utf-8]\n    pretrained_vectors\t   pretrained word vectors (.vec file) for supervised learning []\n\nExample usage:\n\n.. code:: python\n\n    classifier = fasttext.supervised('train.txt', 'model', label_prefix='__myprefix__',\n                                     thread=4)\n\nLoad pre-trained classifier\n~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\nFile ``.bin`` that previously trained or generated by fastText can be\nloaded using this function.\n\n.. code:: shell\n\n    ./fasttext supervised -input train.txt -output classifier -label 'some_prefix'\n\n.. code:: python\n\n    classifier = fasttext.load_model('classifier.bin', label_prefix='some_prefix')\n\nTest classifier\n~~~~~~~~~~~~~~~\n\nThis is equivalent as ``fasttext(1)`` test command. The test using the\nsame model and test set will produce the same value for the precision at\none and the number of examples.\n\n.. code:: python\n\n    result = classifier.test(params)\n\n    # Properties\n    result.precision # Precision at one\n    result.recall    # Recall at one\n    result.nexamples # Number of test examples\n\nThe param ``k`` is optional, and equal to ``1`` by default.\n\nPredict the most-likely label of texts\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\nThis interface is equivalent as ``fasttext(1)`` predict command.\n\n``texts`` is an array of string\n\n.. code:: python\n\n    labels = classifier.predict(texts, k)\n\n    # Or with probability\n    labels = classifier.predict_proba(texts, k)\n\nThe param ``k`` is optional, and equal to ``1`` by default.\n\nAttributes and methods for the classifier\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\nClassifier have the following atributes & methods\n\n.. code:: python\n\n    classifier.labels                  # List of labels\n    classifier.label_prefix            # Prefix of the label\n    classifier.dim                     # Size of word vector\n    classifier.ws                      # Size of context window\n    classifier.epoch                   # Number of epochs\n    classifier.min_count               # Minimal number of word occurences\n    classifier.neg                     # Number of negative sampled\n    classifier.word_ngrams             # Max length of word ngram\n    classifier.loss_name               # Loss function name\n    classifier.bucket                  # Number of buckets\n    classifier.minn                    # Min length of char ngram\n    classifier.maxn                    # Max length of char ngram\n    classifier.lr_update_rate          # Rate of updates for the learning rate\n    classifier.t                       # Value of sampling threshold\n    classifier.encoding                # Encoding that used by classifier\n    classifier.test(filename, k)       # Test the classifier\n    classifier.predict(texts, k)       # Predict the most likely label\n    classifier.predict_proba(texts, k) # Predict the most likely label include their probability\n\nThe param ``k`` for ``classifier.test``, ``classifier.predict`` and\n``classifier.predict_proba`` is optional, and equal to ``1`` by default.\n\nReferences\n----------\n\nEnriching Word Vectors with Subword Information\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\n[1] P. Bojanowski\\*, E. Grave\\*, A. Joulin, T. Mikolov, `*Enriching Word\nVectors with Subword\nInformation* <https://arxiv.org/pdf/1607.04606v1.pdf>`__\n\n::\n\n    @article{bojanowski2016enriching,\n      title={Enriching Word Vectors with Subword Information},\n      author={Bojanowski, Piotr and Grave, Edouard and Joulin, Armand and Mikolov, Tomas},\n      journal={arXiv preprint arXiv:1607.04606},\n      year={2016}\n    }\n\nBag of Tricks for Efficient Text Classification\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\n[2] A. Joulin, E. Grave, P. Bojanowski, T. Mikolov, `*Bag of Tricks for\nEfficient Text\nClassification* <https://arxiv.org/pdf/1607.01759v2.pdf>`__\n\n::\n\n    @article{joulin2016bag,\n      title={Bag of Tricks for Efficient Text Classification},\n      author={Joulin, Armand and Grave, Edouard and Bojanowski, Piotr and Mikolov, Tomas},\n      journal={arXiv preprint arXiv:1607.01759},\n      year={2016}\n    }\n\n(\\* These authors contributed equally.)\n\nJoin the fastText community\n---------------------------\n\n-  Facebook page: https://www.facebook.com/groups/1174547215919768\n-  Google group:\n   https://groups.google.com/forum/#!forum/fasttext-library\n\n.. |Build Status| image:: https://travis-ci.org/salestock/fastText.py.svg?branch=master\n   :target: https://travis-ci.org/salestock/fastText.py\n.. |PyPI version| image:: https://badge.fury.io/py/fasttext.svg\n   :target: https://badge.fury.io/py/fasttext\n\n\n",
        "description_content_type": "",
        "docs_url": null,
        "download_url": "",
        "downloads": {
            "last_day": -1,
            "last_month": -1,
            "last_week": -1
        },
        "home_page": "https://github.com/pyk/fastText.py",
        "keywords": "",
        "license": "BSD 3-Clause License",
        "maintainer": "",
        "maintainer_email": "",
        "name": "fasttext-win",
        "package_url": "https://pypi.org/project/fasttext-win/",
        "platform": "",
        "project_url": "https://pypi.org/project/fasttext-win/",
        "project_urls": {
            "Homepage": "https://github.com/pyk/fastText.py"
        },
        "release_url": "https://pypi.org/project/fasttext-win/0.8.3/",
        "requires_dist": [
            "numpy (>=1)",
            "future"
        ],
        "requires_python": "",
        "summary": "A Python interface for Facebook fastText library",
        "version": "0.8.3"
    },
    "last_serial": 4051749,
    "releases": {
        "0.8.3": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "7af2711d168921023be5131cff8aab85",
                    "sha256": "ba510d486565b8518f258b676c30ef0f502113f4610863ab9a9af9cc954665e4"
                },
                "downloads": -1,
                "filename": "fasttext_win-0.8.3-cp36-cp36m-win_amd64.whl",
                "has_sig": false,
                "md5_digest": "7af2711d168921023be5131cff8aab85",
                "packagetype": "bdist_wheel",
                "python_version": "cp36",
                "requires_python": null,
                "size": 106819,
                "upload_time": "2018-07-11T17:32:41",
                "url": "https://files.pythonhosted.org/packages/10/dd/c2765ebd81fc54dcd22dafb3a818c6e4dce405cbb1174645275193654e18/fasttext_win-0.8.3-cp36-cp36m-win_amd64.whl"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "82c042579a578cc60933763ccf14a1dd",
                    "sha256": "6c65751f1306e42809e15d98f04480a2471f5421f938db71f8e7b3c8ee930255"
                },
                "downloads": -1,
                "filename": "fasttext_win-0.8.3.tar.gz",
                "has_sig": false,
                "md5_digest": "82c042579a578cc60933763ccf14a1dd",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 82316,
                "upload_time": "2018-07-11T17:32:42",
                "url": "https://files.pythonhosted.org/packages/a0/32/000658b6c380a987afdf1c7efee163d2014f18de450bffaca5f247ae8570/fasttext_win-0.8.3.tar.gz"
            }
        ]
    },
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "md5": "7af2711d168921023be5131cff8aab85",
                "sha256": "ba510d486565b8518f258b676c30ef0f502113f4610863ab9a9af9cc954665e4"
            },
            "downloads": -1,
            "filename": "fasttext_win-0.8.3-cp36-cp36m-win_amd64.whl",
            "has_sig": false,
            "md5_digest": "7af2711d168921023be5131cff8aab85",
            "packagetype": "bdist_wheel",
            "python_version": "cp36",
            "requires_python": null,
            "size": 106819,
            "upload_time": "2018-07-11T17:32:41",
            "url": "https://files.pythonhosted.org/packages/10/dd/c2765ebd81fc54dcd22dafb3a818c6e4dce405cbb1174645275193654e18/fasttext_win-0.8.3-cp36-cp36m-win_amd64.whl"
        },
        {
            "comment_text": "",
            "digests": {
                "md5": "82c042579a578cc60933763ccf14a1dd",
                "sha256": "6c65751f1306e42809e15d98f04480a2471f5421f938db71f8e7b3c8ee930255"
            },
            "downloads": -1,
            "filename": "fasttext_win-0.8.3.tar.gz",
            "has_sig": false,
            "md5_digest": "82c042579a578cc60933763ccf14a1dd",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": null,
            "size": 82316,
            "upload_time": "2018-07-11T17:32:42",
            "url": "https://files.pythonhosted.org/packages/a0/32/000658b6c380a987afdf1c7efee163d2014f18de450bffaca5f247ae8570/fasttext_win-0.8.3.tar.gz"
        }
    ]
}