{"id":460,"date":"2018-01-30T05:36:11","date_gmt":"2018-01-30T05:36:11","guid":{"rendered":"http:\/\/bskog.com\/ai\/?p=460"},"modified":"2018-01-30T05:37:24","modified_gmt":"2018-01-30T05:37:24","slug":"fast-ai-2018-released-truly-awesome","status":"publish","type":"post","link":"http:\/\/bskog.com\/ai\/2018\/01\/30\/fast-ai-2018-released-truly-awesome\/","title":{"rendered":"Fast.ai 2018 has been released, and it&#8217;s truly awesome"},"content":{"rendered":"<p>Last year the 2017 course of fast.ai was amazing, which taught state of the art deep learning to coders. There are so many goodies in the <a href=\"http:\/\/www.fast.ai\/2018\/01\/26\/v2-launch\/\">blog post about the Fast.ai 2018 launch\u00a0<\/a>which is available now. This year they held the course using Pytorch instead of Keras and wrote their own library for speeding up development and were the\u00a0first to add several implementations from papers to the library such as\u00a0Learning Rate Finder (<a href=\"https:\/\/arxiv.org\/abs\/1506.01186\">Smith 2015<\/a>) and\u00a0Stochastic Gradient Descent with Restarts (SGDR). With one line of code, you can also get the images that the classifier gets wrong.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-461\" src=\"http:\/\/bskog.com\/ai\/wp-content\/uploads\/sites\/2\/2018\/01\/sgdr.png\" alt=\"\" width=\"977\" height=\"750\" srcset=\"http:\/\/bskog.com\/ai\/wp-content\/uploads\/sites\/2\/2018\/01\/sgdr.png 977w, http:\/\/bskog.com\/ai\/wp-content\/uploads\/sites\/2\/2018\/01\/sgdr-300x230.png 300w, http:\/\/bskog.com\/ai\/wp-content\/uploads\/sites\/2\/2018\/01\/sgdr-768x590.png 768w\" sizes=\"auto, (max-width: 767px) 89vw, (max-width: 1000px) 54vw, (max-width: 1071px) 543px, 580px\" \/> <img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-462\" src=\"http:\/\/bskog.com\/ai\/wp-content\/uploads\/sites\/2\/2018\/01\/lr_finder.png\" alt=\"\" width=\"960\" height=\"753\" srcset=\"http:\/\/bskog.com\/ai\/wp-content\/uploads\/sites\/2\/2018\/01\/lr_finder.png 960w, http:\/\/bskog.com\/ai\/wp-content\/uploads\/sites\/2\/2018\/01\/lr_finder-300x235.png 300w, http:\/\/bskog.com\/ai\/wp-content\/uploads\/sites\/2\/2018\/01\/lr_finder-768x602.png 768w\" sizes=\"auto, (max-width: 767px) 89vw, (max-width: 1000px) 54vw, (max-width: 1071px) 543px, 580px\" \/><\/p>\n<p>17 of the 20 top participants in a\u00a0kaggle\u00a0competitors were students in the preview course.<\/p>\n<p>I\u00a0recommend reading the <a href=\"http:\/\/www.fast.ai\/2018\/01\/26\/v2-launch\/\">blog post<\/a> and taking the course.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Last year the 2017 course of fast.ai was amazing, which taught state of the art deep learning to coders. There are so many goodies in the blog post about the Fast.ai 2018 launch\u00a0which is available now. This year they held the course using Pytorch instead of Keras and wrote their own library for speeding up &hellip; <\/p>\n<p class=\"link-more\"><a href=\"http:\/\/bskog.com\/ai\/2018\/01\/30\/fast-ai-2018-released-truly-awesome\/\" class=\"more-link\">Continue reading<span class=\"screen-reader-text\"> &#8220;Fast.ai 2018 has been released, and it&#8217;s truly awesome&#8221;<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[9,11,13,16],"tags":[],"class_list":["post-460","post","type-post","status-publish","format-standard","hentry","category-courses","category-deep-learning","category-frameworks","category-learning-material"],"_links":{"self":[{"href":"http:\/\/bskog.com\/ai\/wp-json\/wp\/v2\/posts\/460","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/bskog.com\/ai\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/bskog.com\/ai\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/bskog.com\/ai\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/bskog.com\/ai\/wp-json\/wp\/v2\/comments?post=460"}],"version-history":[{"count":2,"href":"http:\/\/bskog.com\/ai\/wp-json\/wp\/v2\/posts\/460\/revisions"}],"predecessor-version":[{"id":464,"href":"http:\/\/bskog.com\/ai\/wp-json\/wp\/v2\/posts\/460\/revisions\/464"}],"wp:attachment":[{"href":"http:\/\/bskog.com\/ai\/wp-json\/wp\/v2\/media?parent=460"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/bskog.com\/ai\/wp-json\/wp\/v2\/categories?post=460"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/bskog.com\/ai\/wp-json\/wp\/v2\/tags?post=460"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}