:orphan:

|rss_image|  **2018-10 - 1/1** :ref:`Blog <ap-main-0>` 

.. |rss_image| image:: feed-icon-16x16.png
    :target: ../_downloads/rss.xml
    :alt: RSS

----


.. index:: 2018-10


.. _ap-month-2018-10-0:

2018-10 - 1/1
+++++++++++++

.. blogpostagg::
    :title: Deep model from one framework to another
    :date: 2018-10-24
    :keywords: deep learning,conversion
    :categories: modules
    :rawfile: 2018/2018-10-24_mdnn.rst

    I discovered the python module
    `MMdnn <https://github.com/Microsoft/MMdnn>`_
    which implements many conversions between frameworks,
    from :epkg:`tensorflow` to :epkg:`pytorch` and many others
    (see `tested models <https://github.com/Microsoft/MMdnn#tested-models>`_).
    It does propose a visualizer too:
    `MMDNN Visualizer <http://mmdnn.eastasia.cloudapp.azure.com:8080/>`_.
    And many models can be downloaded:
    `Model Collection <https://github.com/Microsoft/MMdnn/blob/master/mmdnn/models/README.md#model-collection>`_.
    Available on `PyPi <https://pypi.org/project/mmdnn/#description>`_.





----

|rss_image|  **2018-10 - 1/1** :ref:`2020-10 (1) <ap-month-2020-10-0>` :ref:`2021-01 (1) <ap-month-2021-01-0>` :ref:`2021-06 (2) <ap-month-2021-06-0>` :ref:`2021-11 (1) <ap-month-2021-11-0>` :ref:`2022-02 (1) <ap-month-2022-02-0>`