0

Blocks

一个用于构建和训练神经网络的 Theano 框架。「A Theano framework for building and training neural networks」

Github repo
Main metrics
  • Owner: mila-iqia/blocks
  • Platform: Linux, Mac, Web browsers, Windows
  • License:: Other
  • Category::
  • Topic:
  • Like:
    0
      Compare:

Github stars Tracking Chart

Blocks

Blocks 是一个帮助你在 Theano 之上构建神经网络模型的框架。目前它支持并提供：

  • 构建参数化的 Theano 操作，称为 "bricks"。
  • 在大型模型中选择变量和 bricks 的模式匹配。
  • 优化模型的算法。
  • 训练的保存和恢复。
  • 在训练过程中监测和分析数值（在训练集和测试集上）。
  • 图形变换的应用，如 dropout 等。

在未来，我们也希望能够支持：

  • 尺寸、类型和轴检查

另见

  • Fuel，主要为 Blocks 开发的数据处理引擎。
  • Blocks-examples，用于维护使用 Blocks 的脚本示例。
  • Blocks-extras：半维护的附加 Blocks 组件。

引用Blocks

如果您在工作中使用了 Blocks 或 Fuel，如果您能引用以下论文，我们将非常感激：

Bart van Merriënboer, Dzmitry Bahdanau, Vincent Dumoulin, Dmitriy Serdyuk, David Warde-Farley, Jan Chorowski, and Yoshua Bengio, "Blocks and Fuel: Frameworks for deep learning," arXiv preprint arXiv:1506.00619 [cs.LG], 2015.

文档介绍

更多信息请参见文档。

贡献

如果你想做出贡献，请务必阅读开发者指南。


Main metrics

Overview
Name With Ownermila-iqia/blocks
Primary LanguagePython
Program languagePython (Language Count: 1)
PlatformLinux, Mac, Web browsers, Windows
License:Other
所有者活动
Created At2014-10-06 00:08:32
Pushed At2019-02-19 12:41:38
Last Commit At2018-09-06 12:02:54
Release Count4
Last Release Name0.2 (Posted on )
First Release Namev0.0.1 (Posted on )
用户参与
Stargazers Count1.2k
Watchers Count84
Fork Count348
Commits Count3.3k
Has Issues Enabled
Issues Count535
Issue Open Count183
Pull Requests Count545
Pull Requests Open Count9
Pull Requests Close Count113
项目设置
Has Wiki Enabled
Is Archived
Is Fork
Is Locked
Is Mirror
Is Private

.. image:: https://img.shields.io/coveralls/mila-udem/blocks.svg
:target: https://coveralls.io/r/mila-udem/blocks

.. image:: https://travis-ci.org/mila-udem/blocks.svg?branch=master
:target: https://travis-ci.org/mila-udem/blocks

.. image:: https://readthedocs.org/projects/blocks/badge/?version=latest
:target: https://blocks.readthedocs.org/

.. image:: https://img.shields.io/scrutinizer/g/mila-udem/blocks.svg
:target: https://scrutinizer-ci.com/g/mila-udem/blocks/

.. image:: https://requires.io/github/mila-udem/blocks/requirements.svg?branch=master
:target: https://requires.io/github/mila-udem/blocks/requirements/?branch=master

.. image:: https://img.shields.io/badge/license-MIT-blue.svg
:target: https://github.com/mila-udem/blocks/blob/master/LICENSE

Blocks

Blocks is a framework that helps you build neural network models on top of
Theano. Currently it supports and provides:

  • Constructing parametrized Theano operations, called "bricks"
  • Pattern matching to select variables and bricks in large models
  • Algorithms to optimize your model
  • Saving and resuming of training
  • Monitoring and analyzing values during training progress (on the training set
    as well as on test sets)
  • Application of graph transformations, such as dropout

In the future we also hope to support:

  • Dimension, type and axes-checking

See Also:

  • Fuel_, the data processing engine developed primarily for Blocks.
  • Blocks-examples_ for maintained examples of scripts using Blocks.
  • Blocks-extras_ for semi-maintained additional Blocks components.

Citing Blocks
If you use Blocks or Fuel in your work, we'd really appreciate it if you could cite the following paper:

Bart van Merriënboer, Dzmitry Bahdanau, Vincent Dumoulin, Dmitriy Serdyuk, David Warde-Farley, Jan Chorowski, and Yoshua Bengio, "Blocks and Fuel: Frameworks for deep learning_," arXiv preprint arXiv:1506.00619 [cs.LG], 2015.

Documentation
Please see the documentation_ for more information.

Contributing
If you want to contribute, please make sure to read the developer guidelines_.

.. _documentation: http://blocks.readthedocs.org
.. _developer guidelines: http://blocks.readthedocs.org/en/latest/development/index.html
.. _Blocks and Fuel: Frameworks for deep learning: http://arxiv.org/abs/1506.00619
.. _Blocks-examples: https://github.com/mila-udem/blocks-examples
.. _Blocks-extras: https://github.com/mila-udem/blocks-extras
.. _Fuel: https://github.com/mila-udem/fuel