8个用于深度学习的开源工具

{"type":"doc","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"人工智能是一个正在迅速发展的领域，并日益成为一种主流。要想跟上人工智能技术的发展，最好的方法就是学习深度学习中的新技术。以下这 8 个开源工具，可以帮助你轻松实现深度学习。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"1.TensorFlow"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"这个强大的工具包开发于 2015 年，被誉为机器\/深度学习中最容易使用和部署的工具之一。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"link","attrs":{"href":"https:\/\/www.tensorflow.org\/","title":null,"type":null},"content":[{"type":"text","text":"TensorFlow"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":" 最初是由谷歌大脑团队创建的，用于处理其研究和生产目标，该项目构建了深度神经网络来执行自然语言处理、图像识别和翻译等任务。由于它提供了大量的免费工具、库和社区资源，现在它的使用甚至被 Uber、Twitter 和 eBay 等公司广泛接受。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"它可用于 Python、"},{"type":"link","attrs":{"href":"https:\/\/www.haskell.org\/","title":null,"type":null},"content":[{"type":"text","text":"Haskell"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"、C++、Java 甚至是 JavaScript。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"2.Keras"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"link","attrs":{"href":"https:\/\/keras.io\/","title":null,"type":null},"content":[{"type":"text","text":"Keras"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":" 由Python 编写，最初发布于 2015 年，是一种高级神经网络 "},{"type":"link","attrs":{"href":"https:\/\/www.altexsoft.com\/blog\/engineering\/what-is-api-definition-types-specifications-documentation\/","title":null,"type":null},"content":[{"type":"text","text":"API"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"，旨在简化机器学习和深度学习，可以在 TensorFlow 或 Theano 之上部署。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"Keras 以其高度的广泛性、模块化、易用的特点，通过简单的原型设计实现了快速的实验，在 "},{"type":"link","attrs":{"href":"https:\/\/en.wikipedia.org\/wiki\/Central_processing_unit","title":null,"type":null},"content":[{"type":"text","text":"CPU"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":" 和 "},{"type":"link","attrs":{"href":"https:\/\/en.wikipedia.org\/wiki\/Graphics_processing_unit","title":null,"type":null},"content":[{"type":"text","text":"GPU"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":" 上高效运行，这对研究工作至关重要。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"3.PaddlePaddle"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"并行分布式深度学习，又称 "},{"type":"link","attrs":{"href":"https:\/\/github.com\/PaddlePaddle\/","title":null,"type":null},"content":[{"type":"text","text":"PaddlePaddle"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"，"},{"type":"text","text":"具有易用、高效、灵活和可伸缩等特点"},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"这是百度研发的深度学习平台，为百度内部多项产品提供深度学习算法支持。"},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"它在 2016 年向专业社区开源，具有深度学习的先进功能、端到端的开发工具包，受到制造业和农业部门的使用者青睐。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"4.PyTorch"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"这个开源工具包"},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"使用 Python 脚本语言，一般用于 "},{"type":"link","attrs":{"href":"https:\/\/www.ibm.com\/cloud\/learn\/natural-language-processing#:~:text=Natural%20language%20processing%20(NLP","title":null,"type":null},"content":[{"type":"text","text":"自然语言处理"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"和"},{"type":"link","attrs":{"href":"https:\/\/www.ibm.com\/topics\/computer-vision","title":null,"type":null},"content":[{"type":"text","text":"计算机视觉"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"。它具有强大的 GPU、内存使用效率和动态计算图，这使得它在协助开发 "},{"type":"link","attrs":{"href":"https:\/\/www.mathworks.com\/help\/deeplearning\/ug\/how-dynamic-neural-networks-work.html;jsessionid=0dd555d070bf6f6fd6c6182ca137","title":null,"type":null},"content":[{"type":"text","text":"动态神经网络"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"方面很受欢迎，并能够根据用户的要求建立图形和可视化。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"5.Theano"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"这个开源库"},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"发布于 2007 年，使用 Python 编写脚本，允许用户定义、定制和评估数学表达式，从而使深度学习模型的形成变得容易。由于它是最古老的可自由使用的库之一，它激发了深度学习的发展，导致了高效的"},{"type":"link","attrs":{"href":"https:\/\/en.wikipedia.org\/wiki\/Computer_algebra","title":null,"type":null},"content":[{"type":"text","text":"符号微分"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"（symbolic differentiation）。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"6.Weka"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"link","attrs":{"href":"https:\/\/www.cs.waikato.ac.nz\/ml\/weka\/","title":null,"type":null},"content":[{"type":"text","text":"Weka"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","text":" "},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"最初是作为一个机器学习开源工具包开发的，它还具有深度学习包的功能，无论是在包管理、分类还是可视化方面，它都可以在一个简单的界面使用中进行更复杂的分析。它是一个 Java 应用程序，但可以通过"},{"type":"link","attrs":{"href":"https:\/\/www.ibm.com\/cloud\/learn\/api#:~:text=Application%20programming%20interfaces%2C%20or%20APIs,and%20functionality%20easily%20and%20securely.","title":null,"type":null},"content":[{"type":"text","text":"应用编程接口"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"用于 R 或 Python。因为它最初是为了一个教学平台而开发的，一般只限于教学工具或者小项目。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"7.Caffe"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"link","attrs":{"href":"http:\/\/caffe.berkeleyvision.org\/","title":null,"type":null},"content":[{"type":"text","text":"Caffe"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"（Convolutional Architecture for Fast Feature Embedding，意即快速特征嵌入的卷积架构）是一个开源的深度学习网络，是为速度、表达和模块化设计的。它使用了 "},{"type":"link","attrs":{"href":"https:\/\/en.wikipedia.org\/wiki\/C%2B%2B","title":null,"type":null},"content":[{"type":"text","text":"C++"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"，但也有一个 Python 接口。它由伯克利人工智能研究团队于 2017 年开发并发布，它有一个精心设计的架构，良好的代码编写和快速的性能，可以快速实现工业部署。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"9.Torch"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"link","attrs":{"href":"http:\/\/torch.ch\/","title":null,"type":null},"content":[{"type":"text","text":"Torch"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","text":" "},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":"最初是在 2002 年作为机器学习库开发和发布的，它提供了一系列用于深度学习的算法，重点是 GPU，并提供 "},{"type":"link","attrs":{"href":"https:\/\/en.wikipedia.org\/wiki\/IOS","title":null,"type":null},"content":[{"type":"text","text":"iOS"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":" 和 "},{"type":"link","attrs":{"href":"https:\/\/www.techopedia.com\/definition\/4219\/android-platform","title":null,"type":null},"content":[{"type":"text","text":"Android"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":" 平台支持。它由脚本语言 "},{"type":"link","attrs":{"href":"https:\/\/www.lua.org\/about.html","title":null,"type":null},"content":[{"type":"text","text":"LUA"}],"marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}]},{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":" 和底层的 C 组件组成，使得它使用起来简单、高效、快速。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"color","attrs":{"color":"#494949","name":"user"}}],"text":" "}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"作者介绍："}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Rancho Labs，科技爱好者自媒体，培养年轻人的思维，使他们在编程\/机器人技术\/人工智能领域站稳脚跟。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"原文链接："}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"https:\/\/rancholabs.medium.com\/open-source-tools-for-deep-learning-f4d68c2a39cd"}]}]}