Torch vs Theano

Recently we took a look at Torch 7 and found its data ingestion facilities less than impressive. Torch’s biggest competitor seems to be Theano, a popular deep-learning framework for Python.

It seems that these two have been having “who is faster” competition going for a few years now. It’s been documented in the following papers:

  1. J. Bergstra, O. Breuleux, F. Bastien, P. Lamblin, R. Pascanu, G. Desjardins, J. Turian, Y. Bengio - Theano: a CPU and GPU Math Expression Compiler PDF

  2. Ronan Collobert, Koray Kavukcuoglu, Clement Farabet - Torch7: A Matlab-like Environment for Machine Learning PDF

  3. Frédéric Bastien, Pascal Lamblin, Razvan Pascanu, James Bergstra, Ian Goodfellow, Arnaud Bergeron, Nicolas Bouchard, David Warde-Farley, Yoshua Bengio - Theano: new features and speed improvements arxiv


A figure from the Torch7 paper [2]. Torch - red, Theano - green. Higher is better.

And a quote from [3]:

Bergstra et al.(2010) showed that Theano was faster than many other tools available at the time, including Torch5. The following year, Collobert et al.(2011) showed that Torch7 was faster than Theano on the same benchmarks.

The results in the last paper are mixed, if you’re wondering.

The latest act in this friendly competition, which can be seen as one between Bengio’s and LeCun’s groups, appears to be about FFT convolutions, first available in Theano and recently open-sourced by Facebook in Torch.

As a side note, the press really jumped at this second event with headlines about turbo-charging deep learning and the like. Probably the allure of Facebook and deep learning in the same sentence.

Let’s look at convnet benchmarks by Soumith Chintala. He is a Facebook/Torch guy and yet the Theano’s convolution layer is reported to be the fastest at the time of writing. Waiting for those fbfft results.

Anyway, speed isn’t everything and there’s more to life than FFT convolutions. From a developer’s perspective minor differences in speed are less important than other factors, like ease of use. Which leads us to what Soumith had to say about Torch, according to VentureBeat:

It’s like building some kind of electronic contraption or, like, a Lego set. You just can plug in and plug out all these blocks that have different dynamics and that have complex algorithms within them.

At the same time Torch is actually not extremely difficult to learn — unlike, say, the Theano library.

We’ve made it incredibly easy to use. We introduce someone to Torch, and they start churning out research really fast.

Well, you already know our opinion about the “incredibly easy” bit. Torch is not really a Matlab-like environment. Matlab, with all its shortcomings, is a very well polished piece of software with examplary documentation. Torch, on the other hand, is rather rough around the edges.

Besides the language gap, that’s one of the reasons that you don’t see that much Torch usage apart from Facebook and DeepMind. At the same time libraries using Theano have been springing up like mushrooms after a rain (you might want to take a look at Sander Dieleman’s Lasagne and at blocks). It is hard to beat the familiar and rich Python ecosystem.

Theano tutorials

P.S. What about Caffe?

Caffe is a fine and very popular piece of software. How does it compare with Torch and Theano? Here’s sieisteinmodel’s answer from Reddit:

Caffe has a pretty different target. More mass market, for people who want to use deep learning for applications. Torch and Theano are more tailored towards people who want to use it for research on DL itself.

Torch vs Theano的更多相关文章

  1. mxnet,theano与torch的简单比较

    这篇文章我想来比较一下Theano和mxnet,Torch(Torch基本没用过,所以只能说一些直观的感觉).我主要从以下几个方面来计较它们: 1.学习框架的成本,接口设计等易用性方面. 三个框架的学 ...

  2. Summary on deep learning framework --- Theano && Lasagne

     Summary on deep learning framework --- Theano && Lasagne 2017-03-23 1. theano.function outp ...

  3. 普通程序员如何转向AI方向

    眼下,人工智能已经成为越来越火的一个方向.普通程序员,如何转向人工智能方向,是知乎上的一个问题.本文是我对此问题的一个回答的归档版.相比原回答有所内容增加. 一. 目的 本文的目的是给出一个简单的,平 ...

  4. AI方向

    普通程序员如何转向AI方向   眼下,人工智能已经成为越来越火的一个方向.普通程序员,如何转向人工智能方向,是知乎上的一个问题.本文是我对此问题的一个回答的归档版.相比原回答有所内容增加. 一. 目的 ...

  5. (转) Deep Learning Resources

    转自:http://www.jeremydjacksonphd.com/category/deep-learning/ Deep Learning Resources Posted on May 13 ...

  6. 学习Data Science/Deep Learning的一些材料

    原文发布于我的微信公众号: GeekArtT. 从CFA到如今的Data Science/Deep Learning的学习已经有一年的时间了.期间经历了自我的兴趣.擅长事务的探索和试验,有放弃了的项目 ...

  7. 百度Paddle会和Python一样,成为最流行的深度学习引擎吗?

    PaddlePaddle会和Python一样流行吗? 深度学习引擎最近经历了开源热.2013年Caffe开源,很快成为了深度学习在图像处理中的主要框架,但那时候的开源框架还不多.随着越来越多的开发者开 ...

  8. Google研究员Ilya Sutskever:成功训练LDNN的13点建议

    Google研究员Ilya Sutskever:成功训练LDNN的13点建议 摘要:本文由Ilya Sutskever(Google研究员.深度学习泰斗Geoffrey Hinton的学生.DNNre ...

  9. Popular Deep Learning Tools – a review

    Popular Deep Learning Tools – a review Deep Learning is the hottest trend now in AI and Machine Lear ...

随机推荐

  1. Iperf使用方法

    Iperf使用方法 Iperf  是一个网络性能测试工具.Iperf可以测试TCP和UDP带宽质量.Iperf可以测量最大TCP带宽,具有多种参数和UDP特性.Iperf可以报告带宽,延迟抖动和数据包 ...

  2. KALI 2.0优化

    1.首选项没有主菜单,系统管理 apt-get install alacarte 2.APT源 #官方源deb http://http.kali.org/kali kali main non-free ...

  3. JQuery设置input属性(disabled、enabled)

    document.getElementById("removeButton").disabled = false; //普通Js写法 $("#removeButton&q ...

  4. C# Mutex对象的使用

    C# Mutex对象的使用 C#语言有很多值得学习的地方,这里我们主要介绍C# Mutex对象,包括介绍控制好多个线程相互之间的联系等方面. 如何控制好多个线程相互之间的联系,不产生冲突和重复,这需要 ...

  5. Service解析

    Service解析: 运行service有如下两种方式: StartService() 访问者退出,service仍然运行: BindService() 访问者与service绑定,访问者退出,ser ...

  6. php注册登录时生成的验证码

    http://blog.sina.com.cn/s/blog_8173443e01012l82.html 记得我学php时第一件事就是研究登陆注册.当然,登陆少不了验证码.两年过去了,昨天突然想用个验 ...

  7. Candence下对“跨页连接器(off-page connector)”进行批量重命名的方法

    parts.ports.alias等等均可以在“属性编辑器(Property Editor)”中进行查看编辑,并通过复制到Excel等表格软件来进行批量修改.之后再粘贴回去的方法进行批量编辑.但是“跨 ...

  8. Solr使用初探——Solr的安装环境与配置

    Solr是一个apache名下很好用的开源索引.搜索工具,网上的资料虽多但很杂,笔者花了一天的时间对Solr进行了较为初步的研究,对Solr的基础应用做了一定的总结.文中涉及到的配置方法并不唯一,AP ...

  9. oracle11g不能导出空表

    Oracle11g的数据库迁移,习惯性的用了exp/imp,然后在新的数据库发现,空表根本没有exp出来,然后查资料,发现了如下信息: ORACLE 11G在用EXPORT导出时,空表不能导出. 11 ...

  10. ios应用来电监听

    先导入这两个头文件,库文件不用导可以 #import <CoreTelephony/CTCallCenter.h> #import <CoreTelephony/CTCall.h&g ...