{
    "info": {
        "author": "Danial Alihosseini",
        "author_email": "danial.alihosseini@gmail.com",
        "bugtrack_url": null,
        "classifiers": [
            "Development Status :: 3 - Alpha",
            "License :: OSI Approved :: MIT License",
            "Operating System :: POSIX :: Linux",
            "Programming Language :: Python :: 3"
        ],
        "description": "# FastBLEU Package\n\nThis is a fast multithreaded C++ implementation of NLTK BLEU; computing BLEU and SelfBLEU score for a fixed reference set.\nIt can return (Self)BLEU for different (max) n-grams simultaneously and efficiently (e.g. BLEU-2, BLEU-3 and etc.).\n\n## Installation\nPyPI latest stable release\n\n~~~~~~~~~~~~~~~~~~~~~~~~\npip install --user FastBLEU\n~~~~~~~~~~~~~~~~~~~~~~~~\n\n## Sample Usage\nHere is an example to compute BLEU-2, BLEU-3, SelfBLEU-2 and SelfBLEU-3:\n\n```python\n>>> from fast_bleu import BLEU, SelfBLEU\n>>> ref1 = ['It', 'is', 'a', 'guide', 'to', 'action', 'that',\n...          'ensures', 'that', 'the', 'military', 'will', 'forever',\n...          'heed', 'Party', 'commands']\n>>> ref2 = ['It', 'is', 'the', 'guiding', 'principle', 'which',\n...          'guarantees', 'the', 'military', 'forces', 'always',\n...          'being', 'under', 'the', 'command', 'of', 'the', 'Party']\n>>> ref3 = ['It', 'is', 'the', 'practical', 'guide', 'for', 'the',\n...          'army', 'always', 'to', 'heed', 'the', 'directions',\n...          'of', 'the', 'party']\n\n>>> hyp1 = ['It', 'is', 'a', 'guide', 'to', 'action', 'which',\n...         'ensures', 'that', 'the', 'military', 'always',\n...         'obeys', 'the', 'commands', 'of', 'the', 'party']\n>>> hyp2 = ['he', 'read', 'the', 'book', 'because', 'he', 'was',\n...         'interested', 'in', 'world', 'history']\n\n>>> list_of_references = [ref1, ref2, ref3]\n>>> hypotheses = [hyp1, hyp2]\n>>> weights = {'bigram': (1/2., 1/2.), 'trigram': (1/3., 1/3., 1/3.)}\n\n>>> bleu = BLEU(list_of_references, weights)\n>>> bleu.get_score(hypotheses)\n{'bigram': [0.7453559924999299, 0.0191380231127159], 'trigram': [0.6240726901657495, 0.013720869575946234]}\n```\n\nwhich means:\n* BLEU-2 for hyp1 is 0.7453559924999299\n* BLEU-2 for hyp2 is 0.0191380231127159\n\n* BLEU-3 for hyp1 is 0.6240726901657495\n* BLEU-3 for hyp2 is 0.013720869575946234\n\n```python\n>>> self_bleu = SelfBLEU(list_of_references, weights)\n>>> self_bleu.get_score()\n{'bigram': [0.25819888974716115, 0.3615507630310936, 0.37080992435478316],\n        'trigram': [0.07808966062765045, 0.20140620205719248, 0.21415334758254043]}\n```\n\nwhich means: \n* SelfBLEU-2 for ref1 is 0.25819888974716115\n* SelfBLEU-2 for ref2 is 0.3615507630310936\n* SelfBLEU-2 for ref3 is 0.37080992435478316\n\n* SelfBLEU-3 for ref1 is 0.07808966062765045\n* SelfBLEU-3 for ref2 is 0.20140620205719248\n* SelfBLEU-3 for ref3 is 0.21415334758254043\n\n**Caution** Each token of reference set is converted to string format during computation.\n\nFor further details, refer to the documentation provided in the source codes.\n\n# Citation\n\nPlease cite our paper if it helps with your research.\n\n* ACL Anthology: https://www.aclweb.org/anthology/W19-2311\n* Arxiv link: https://arxiv.org/abs/1904.03971\n\n```latex\n@article{https://arxiv.org/abs/1904.03971,\n  title={Jointly Measuring Diversity and Quality in Text Generation Models},\n  author={Montahaei, Ehsan and Alihosseini, Danial and Baghshah, Mahdieh Soleymani},\n  journal={NAACL HLT 2019},\n  pages={90},\n  year={2019}\n}\n```",
        "description_content_type": "text/markdown",
        "docs_url": null,
        "download_url": "",
        "downloads": {
            "last_day": -1,
            "last_month": -1,
            "last_week": -1
        },
        "home_page": "https://github.com/Danial-Alh/FastBLEU",
        "keywords": "",
        "license": "OSI Approved :: MIT License",
        "maintainer": "",
        "maintainer_email": "",
        "name": "FastBLEU",
        "package_url": "https://pypi.org/project/FastBLEU/",
        "platform": "POSIX :: Linux",
        "project_url": "https://pypi.org/project/FastBLEU/",
        "project_urls": {
            "Homepage": "https://github.com/Danial-Alh/FastBLEU"
        },
        "release_url": "https://pypi.org/project/FastBLEU/0.0.4/",
        "requires_dist": null,
        "requires_python": ">=3",
        "summary": "A fast multithreaded C++ implementation of nltk BLEU.",
        "version": "0.0.4"
    },
    "last_serial": 5752883,
    "releases": {
        "0.0.1": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "ea0f106c1d500d9a16797d260bf630d6",
                    "sha256": "bf5c81cdca0dc5dd967ad6d4a13da189dd152ba7d58d386a934a867b08808bbb"
                },
                "downloads": -1,
                "filename": "FastBLEU-0.0.1.tar.gz",
                "has_sig": false,
                "md5_digest": "ea0f106c1d500d9a16797d260bf630d6",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": ">=3",
                "size": 7447,
                "upload_time": "2019-08-27T14:58:11",
                "url": "https://files.pythonhosted.org/packages/dc/57/c2fcf06e8ca1ea59c8334562ef92b690d1c2f40757e5405722f4f7af8ca8/FastBLEU-0.0.1.tar.gz"
            }
        ],
        "0.0.3": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "e64480469abdbe6d95ec1356f9b6f3ea",
                    "sha256": "a886361d22ea664b1525f5d839f64b69294a266a708aaf35570a982c54c964cd"
                },
                "downloads": -1,
                "filename": "FastBLEU-0.0.3.tar.gz",
                "has_sig": false,
                "md5_digest": "e64480469abdbe6d95ec1356f9b6f3ea",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": ">=3",
                "size": 9277,
                "upload_time": "2019-08-28T20:59:36",
                "url": "https://files.pythonhosted.org/packages/ab/51/f286a978d6c50cf63249d6034bf776961c21381bb4f947c99573fbdbb2d6/FastBLEU-0.0.3.tar.gz"
            }
        ],
        "0.0.4": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "f96af9f4b2b9fd535fd5a6abb81be4a9",
                    "sha256": "84fd897e1fc6a5f6f1028e386f6ceb1d7ea910ffa4c87045ee3ebc7a6a6c288f"
                },
                "downloads": -1,
                "filename": "FastBLEU-0.0.4.tar.gz",
                "has_sig": false,
                "md5_digest": "f96af9f4b2b9fd535fd5a6abb81be4a9",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": ">=3",
                "size": 9701,
                "upload_time": "2019-08-29T11:40:27",
                "url": "https://files.pythonhosted.org/packages/f3/fb/89a08f6132b66635e8ac29a01e6f35d5b8be75b5d73690b3bf72b7813074/FastBLEU-0.0.4.tar.gz"
            }
        ]
    },
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "md5": "f96af9f4b2b9fd535fd5a6abb81be4a9",
                "sha256": "84fd897e1fc6a5f6f1028e386f6ceb1d7ea910ffa4c87045ee3ebc7a6a6c288f"
            },
            "downloads": -1,
            "filename": "FastBLEU-0.0.4.tar.gz",
            "has_sig": false,
            "md5_digest": "f96af9f4b2b9fd535fd5a6abb81be4a9",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": ">=3",
            "size": 9701,
            "upload_time": "2019-08-29T11:40:27",
            "url": "https://files.pythonhosted.org/packages/f3/fb/89a08f6132b66635e8ac29a01e6f35d5b8be75b5d73690b3bf72b7813074/FastBLEU-0.0.4.tar.gz"
        }
    ]
}