Metadata-Version: 2.0
Name: darkon
Version: 0.0.4
Summary: Performance hacking for your deep learning models
Home-page: https://github.com/darkonhub/darkon
Author: Neosapience, Inc.
Author-email: UNKNOWN
License: Apache License 2.0
Download-URL: https://github.com/darkonhub/darkon/tarball/v0.0.4
Keywords: AI,ML,DL,deep learning,machine learning,neural network,deep neural network,debug neural networks,performance hacking,tensorflow,tf
Platform: UNKNOWN
Classifier: Development Status :: 2 - Pre-Alpha
Classifier: Intended Audience :: Developers
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: Apache Software License
Classifier: Programming Language :: Python :: 2
Classifier: Programming Language :: Python :: 2.7
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.6
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Topic :: Software Development :: Debuggers
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Requires-Dist: numpy
Requires-Dist: six
Requires-Dist: tensorflow (>=1.3.0)

.. role:: raw-html-m2r(raw)
   :format: html



.. raw:: html

   <div align="center">
       <img src="https://cdn.rawgit.com/darkonhub/darkon/d026f574/brand/logo.png" width="400"><br><br>
   </div>



.. image:: https://travis-ci.org/darkonhub/darkon.svg?branch=master
   :target: https://travis-ci.org/darkonhub/darkon
   :alt: Build Status


.. image:: https://codecov.io/gh/darkonhub/darkon/branch/master/graph/badge.svg
   :target: https://codecov.io/gh/darkonhub/darkon
   :alt: codecov


.. image:: https://img.shields.io/badge/License-Apache%202.0-blue.svg
   :target: https://opensource.org/licenses/Apache-2.0
   :alt: License


.. image:: https://img.shields.io/pypi/v/darkon.svg?style=flat-square
   :target: https://pypi.python.org/pypi/darkon
   :alt: PyPI


.. image:: https://badges.gitter.im/darkonhub/darkon.svg
   :target: https://gitter.im/darkonhub/darkon?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge
   :alt: Gitter


.. image:: https://api.codacy.com/project/badge/Grade/077f07f7a52b4d8186beee724ed19231
   :target: https://www.codacy.com/app/zironycho/darkon?utm_source=github.com&amp;utm_medium=referral&amp;utm_content=darkonhub/darkon&amp;utm_campaign=Badge_Grade
   :alt: Codacy Badge


----

**Darkon: Toolkit to Hack Your Deep Learning Models**

**Darkon** is an open source toolkit to understand deep learning models better. Deep learning is often referred as a black-box that is difficult to understand.
But, accountability and controllability could be critical to commercialize deep learning models. People often think that high accuracy on prepared dataset 
is enough to use the model for commercial products. However, well-performing models on prepared dataset often fail in real world usages and cause corner cases 
to be fixed. Moreover, it is necessary to explain the result to trust the system in some applications such as medical diagnosis, financial decisions, etc. We hope\ :raw-html-m2r:`<br>`
**Darkon** can help you to understand the trained models, which could be used to debug failures, interpret decisions, and so on. 

In this first release, we provide influence score calculation easily applicable to any Tensorflow models (other models to be supported later).
The score can be used for filtering bad training samples that affects test performance negatively. 
It can be used for prioritize potential mislabeled examples to be fixed, and debugging distribution mismatch between train and test samples.

We will gradually enable technologies to analyze deep learning models easily applicable to your existing projects.
More features will be released soon. Feedback and feature request are always welcome, which help us to manage priorities. Please keep your eyes on **Darkon**. 

Dependencies
------------


* `Tensorflow <https://github.com/tensorflow/tensorflow>`_\ >=1.3.0

Installation
------------

.. code-block:: bash

   pip install darkon

Usage
-----

.. code-block:: python

   inspector = darkon.Influence(workspace_path,
                                YourDataFeeder(),
                                loss_op_train,
                                loss_op_test,
                                x_placeholder,
                                y_placeholder)

   scores = inspector.upweighting_influence_batch(sess,
                                                  test_indices,
                                                  test_batch_size,
                                                  approx_params,
                                                  train_batch_size,
                                                  train_iterations)

Examples
--------


* `Examples <https://github.com/darkonhub/darkon-examples>`_ 

API Documentation
-----------------


* `Documentation <http://darkon.io/api>`_

Communication
-------------


* `Issues <https://github.com/darkonhub/darkon/issues>`_\ : report issues, bugs, and request new features
* `Pull request <https://github.com/darkonhub/darkon/pulls>`_
* Discuss: `Gitter <https://gitter.im/darkonhub/darkon?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge>`_
* Email: `contact@darkon.io <mailto:contact@darkon.io>`_ 

Authors
-------

`Neosapience, Inc. <http://www.neosapience.com>`_

License
-------

**Apache License 2.0**

References
----------

[1] Cook, R. D. and Weisberg, S. "\ `Residuals and influence in regression <https://www.casact.org/pubs/proceed/proceed94/94123.pdf>`_\ ", New York: Chapman and Hall, 1982

[2] Koh, P. W. and Liang, P. "\ `Understanding Black-box Predictions via Influence Functions <https://arxiv.org/abs/1703.04730>`_\ " ICML2017

[3] Pearlmutter, B. A. "\ `Fast exact multiplication by the hessian <http://www.bcl.hamilton.ie/~barak/papers/nc-hessian.pdf>`_\ " Neural Computation, 1994

[4] Agarwal, N., Bullins, B., and Hazan, E. "\ `Second order stochastic optimization in linear time <https://arxiv.org/abs/1602.03943>`_\ " arXiv preprint arXiv:1602.03943


