{
    "info": {
        "author": "YaHei",
        "author_email": "hey-yahei@qq.com",
        "bugtrack_url": null,
        "classifiers": [
            "Programming Language :: Python :: 3"
        ],
        "description": "## MXOP: MXNet-OpSummary    \nIt only works for **gluon** yet.     \n\nReference: [THOP: PyTorch-OpCounter](https://github.com/Lyken17/pytorch-OpCounter)    \n\n### Installation    \n* PyPi    \n    ```bash\n    pip install --index-url https://pypi.org/simple/ mxop\n    ```\n* Github (latest)    \n    ```bash\n    pip install --upgrade git+https://github.com/hey-yahei/OpSummary.MXNet.git\n    ```\n\n### Usage\n#### Gluon\n* Count OPs    \n    ```python\n    from mxop.gluon import count_ops\n    op_counter = count_ops(net)   # net is the gluon model you want to count OPs \n    ```\n* Count parameters    \n    ```python\n    from mxop.gluon import count_params\n    params_counter = count_params(net, input_size)   # net is the gluon model you want to count parameters\n                                                     # input_size is the shape of your input \n    ```\n* Print summary     \n    ```python\n    from mxop.gluon import op_summary\n    op_summary(net, input_size)   # net is the gluon model you want to count\n                                  # input_size is the shape of your input \n    ```\n\n### Test\n\nRun `tests/test_gluon_utils.py` to count OPs and parameters for all models in model zoo of MXNet.   \n\n#### Result:\n| Model   | Params(M) | Multiplication(G) | \\*Params(M) | *Multiplication(G) | Top1 Acc | Top5 Acc |\n|---:|---:|---:|---:|---:|---:|---:|\n|AlexNet|61.10|0.71|2.47|0.66|0.5492|0.7803|\n|VGG11|132.86|7.61|9.22|7.49|0.6662|0.8734|\n|VGG13|133.04|11.30|9.40|11.18|0.6774|0.8811|\n|VGG16|138.63|15.47|14.71|15.35|0.7323|0.9132|\n|VGG19|143.67|19.63|20.02|19.51|0.7411|0.9135|\n|VGG11_bn|132.87|7.62|9.23|7.49|0.6859|0.8872|\n|VGG13_bn|133.06|11.32|9.42|11.20|0.6884|0.8882|\n|VGG16_bn|138.37|15.48|14.73|15.36|0.7310|0.9176|\n|VGG19_bn|143.69|19.65|20.05|19.52|0.7433|0.9185|\n|Inception_v3|23.87|5.72|21.82|5.72|0.7755|0.9364|\n|ResNet18_v1|11.70|1.82|11.19|1.82|0.7093|0.8992|\n|ResNet34_v1|21.81|3.67|21.3|3.67|0.7437|0.9187|\n|ResNet50_v1|25.63|3.87|23.58|3.87|0.7647|0.9313|\n|ResNet101_v1|44.70|7.59|42.65|7.58|0.7834|0.9401|\n|ResNet152_v1|60.40|11.30|58.36|11.30|0.7900|0.9438|\n|ResNet18_v2|11.70|1.82|11.18|1.82|0.7100|0.8992|\n|ResNet34_v2|21.81|3.67|21.30|3.67|0.7440|0.9208|\n|ResNet50_v2|25.60|4.10|23.55|4.10|0.7711|0.9343|\n|ResNet101_v2|44.64|7.82|42.59|7.81|0.7853|0.9417|\n|ResNet152_v2|60.33|11.54|58.28|11.53|0.7921|0.9431|\n|DenseNet121|8.06|2.85|7.04|2.85|0.7497|0.9225|\n|DenseNet161|28.90|7.76|26.69|7.76|0.7770|0.9380|\n|DenseNet169|14.31|3.38|12.64|3.38|0.7617|0.9317|\n|DenseNet201|20.24|4.32|18.32|4.31|0.7732|0.9362|\n|MobileNet_v1_1.00|4.25|0.57|3.23|0.57|0.7105|0.9006|\n|MobileNet_v1_0.75|2.60|0.33|1.83|0.33|0.6738|0.8782|\n|MobileNet_v1_0.50|1.34|0.15|0.83|0.15|0.6307|0.8475|\n|MobileNet_v1_0.25|0.48|0.04|0.22|0.04|0.5185|0.7608|\n|MobileNet_v2_1.00|3.54|0.32|2.26|0.32|0.7192|0.9056|\n|MobileNet_v2_0.75|2.65|0.19|1.37|0.19|0.6961|0.8895|\n|MobileNet_v2_0.50|1.98|0.10|0.70|0.09|0.6449|0.8547|\n|MobileNet_v2_0.25|1.53|0.03|0.25|0.03|0.5074|0.7456|\n|SqueezeNet1_0|1.25|0.82|0.74|0.73|0.5611|0.7909|\n|SqueezeNet1_1|1.24|0.35|0.72|0.26|0.5496|0.7817|\n\n**To compare for classification models used as backbone--**   \n**\\*Params col shows the number of parameters for models without last several layers.**    \n**\\*Multiplication col shows the number of Multiplication for models without last several layers.**     \n\n![Parameters](http://hey-yahei.cn/imgs/MXNet-OpSummary/Parameters.jpg)\n\n![Multiplication](http://hey-yahei.cn/imgs/MXNet-OpSummary/Multiplication.jpg)   \n\n### TODO\n\n- [ ] Count OPs and parameters for each layer.\n- [ ] Support Symbol model for MXNet.      \n- [ ] Support quantized models.\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": "https://github.com/hey-yahei/OpSummary.MXNet",
        "keywords": "",
        "license": "MIT",
        "maintainer": "",
        "maintainer_email": "",
        "name": "mxop",
        "package_url": "https://pypi.org/project/mxop/",
        "platform": "",
        "project_url": "https://pypi.org/project/mxop/",
        "project_urls": {
            "Homepage": "https://github.com/hey-yahei/OpSummary.MXNet"
        },
        "release_url": "https://pypi.org/project/mxop/0.2.0/",
        "requires_dist": [
            "mxnet"
        ],
        "requires_python": "",
        "summary": "A tool to count OPs and paramters of MXNet model.",
        "version": "0.2.0"
    },
    "last_serial": 4680150,
    "releases": {
        "0.1.0": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "0efe7f94a01b9a2823b4d1b74248bb23",
                    "sha256": "e44c03fc15913d49fb02de9f3f26253adc3584a79146c762a64c7c2ac16a00cd"
                },
                "downloads": -1,
                "filename": "mxop-0.1.0-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "0efe7f94a01b9a2823b4d1b74248bb23",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 4716,
                "upload_time": "2019-01-06T06:34:58",
                "url": "https://files.pythonhosted.org/packages/48/67/25748032d2b14ad8c0fa037baa0ea422b5ef61167d5a6af7a1d915ca74bc/mxop-0.1.0-py3-none-any.whl"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "37a4f064e498d1f649f4e01fd8cfb10a",
                    "sha256": "74492cc3fd08067161d17fb2fa7068e0275fd11cfc72ddba511cdf8e00986949"
                },
                "downloads": -1,
                "filename": "mxop-0.1.0.tar.gz",
                "has_sig": false,
                "md5_digest": "37a4f064e498d1f649f4e01fd8cfb10a",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 3152,
                "upload_time": "2019-01-06T06:35:02",
                "url": "https://files.pythonhosted.org/packages/75/38/458815379bdf12f08c034678e51f1978f41d7a46aa5e2d08ee9850405a1d/mxop-0.1.0.tar.gz"
            }
        ],
        "0.1.1": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "80c0385d0d3124c28a02cf1471447a45",
                    "sha256": "da62995e2425d15ef93e8180de79e7a0cbb09b866bde1f2adfd26db1b36d14fd"
                },
                "downloads": -1,
                "filename": "mxop-0.1.1-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "80c0385d0d3124c28a02cf1471447a45",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 4819,
                "upload_time": "2019-01-06T06:47:34",
                "url": "https://files.pythonhosted.org/packages/d2/51/066073ac58b1a9023066e99b86f7abea3c451c6eb85a865ab0a78903806a/mxop-0.1.1-py3-none-any.whl"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "403c67f983ece876a1c142709ffd6c3d",
                    "sha256": "f84086c8e271d4952735560f0d498ce8c98caa74b7cc3c3a35e0a8f5d24f3f08"
                },
                "downloads": -1,
                "filename": "mxop-0.1.1.tar.gz",
                "has_sig": false,
                "md5_digest": "403c67f983ece876a1c142709ffd6c3d",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 3278,
                "upload_time": "2019-01-06T06:47:35",
                "url": "https://files.pythonhosted.org/packages/1b/c3/d8ce4d58cc970ecce45755a6de593e963115909433872fec77a01a6aeb72/mxop-0.1.1.tar.gz"
            }
        ],
        "0.1.2": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "480f1c16329ee784e183a9337ce09941",
                    "sha256": "266b9600f3ac06930bb17d8462e151e853b4389ebfeac736adcf84cd9b055471"
                },
                "downloads": -1,
                "filename": "mxop-0.1.2-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "480f1c16329ee784e183a9337ce09941",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 4810,
                "upload_time": "2019-01-06T06:57:19",
                "url": "https://files.pythonhosted.org/packages/54/2b/2123dabdebc28cefddb3667f409a1978b91ee5b959bc1175601a77416a36/mxop-0.1.2-py3-none-any.whl"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "c8a1578396d4ac1f21d678a4718829a6",
                    "sha256": "22910304e19c99d8940ba5144cc23c5204af3180822bf514dbfc20f24e756b18"
                },
                "downloads": -1,
                "filename": "mxop-0.1.2.tar.gz",
                "has_sig": false,
                "md5_digest": "c8a1578396d4ac1f21d678a4718829a6",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 3277,
                "upload_time": "2019-01-06T06:57:20",
                "url": "https://files.pythonhosted.org/packages/fb/26/4cfe02834b701158c22290558d135489b9436e9eee2f7bfddb506ae315cb/mxop-0.1.2.tar.gz"
            }
        ],
        "0.2.0": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "b1c969e3d25e5f475a6be058797d610c",
                    "sha256": "3f6752c5f366ab42c5e53526d8001a9b7de5127ad197e6b664e1892bbf68cb8a"
                },
                "downloads": -1,
                "filename": "mxop-0.2.0-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "b1c969e3d25e5f475a6be058797d610c",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 5875,
                "upload_time": "2019-01-10T08:19:32",
                "url": "https://files.pythonhosted.org/packages/89/0b/c15b81c9afb26a780882f6e8691cd96c5552f6701dcefd915886750d9702/mxop-0.2.0-py3-none-any.whl"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "c0648f6a26f7e98a0b0a33c212dc7537",
                    "sha256": "074ce92428cf53b96621a8425295d6fa3fff9b2539b58e102790d34ea630b2a9"
                },
                "downloads": -1,
                "filename": "mxop-0.2.0.tar.gz",
                "has_sig": false,
                "md5_digest": "c0648f6a26f7e98a0b0a33c212dc7537",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 4690,
                "upload_time": "2019-01-10T08:19:34",
                "url": "https://files.pythonhosted.org/packages/e7/4e/cd2728594b6d14113deb15223c912ebf2eb29fe68206efdd782e97f51c82/mxop-0.2.0.tar.gz"
            }
        ]
    },
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "md5": "b1c969e3d25e5f475a6be058797d610c",
                "sha256": "3f6752c5f366ab42c5e53526d8001a9b7de5127ad197e6b664e1892bbf68cb8a"
            },
            "downloads": -1,
            "filename": "mxop-0.2.0-py3-none-any.whl",
            "has_sig": false,
            "md5_digest": "b1c969e3d25e5f475a6be058797d610c",
            "packagetype": "bdist_wheel",
            "python_version": "py3",
            "requires_python": null,
            "size": 5875,
            "upload_time": "2019-01-10T08:19:32",
            "url": "https://files.pythonhosted.org/packages/89/0b/c15b81c9afb26a780882f6e8691cd96c5552f6701dcefd915886750d9702/mxop-0.2.0-py3-none-any.whl"
        },
        {
            "comment_text": "",
            "digests": {
                "md5": "c0648f6a26f7e98a0b0a33c212dc7537",
                "sha256": "074ce92428cf53b96621a8425295d6fa3fff9b2539b58e102790d34ea630b2a9"
            },
            "downloads": -1,
            "filename": "mxop-0.2.0.tar.gz",
            "has_sig": false,
            "md5_digest": "c0648f6a26f7e98a0b0a33c212dc7537",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": null,
            "size": 4690,
            "upload_time": "2019-01-10T08:19:34",
            "url": "https://files.pythonhosted.org/packages/e7/4e/cd2728594b6d14113deb15223c912ebf2eb29fe68206efdd782e97f51c82/mxop-0.2.0.tar.gz"
        }
    ]
}