{
  "id": 12867,
  "title": "Thanks yr, mike1886, authors of CXXNET etc. for sharing your codes and softwares",
  "url": "/competitions/datasciencebowl/discussion/12867",
  "author_name": "SkyLibrary",
  "post_date": "2015-03-17T18:46:10.047000",
  "votes": 1,
  "comment_count": 0,
  "views": 568,
  "content": "<p>I am sure most people consider working on this kaggle project as a opportunity to learn&nbsp;to use deep learning. Open source codes is a great way for us to get started.</p>\n<p>For me, yr's (ChenglongChen) caffe code and mike1886's lasagne/nolearn codes which implemented realtime augmentation really helped me a lot.</p>\n<p>I also find Sander's Galaxy Zoo code, nagadomi's cifar10 code quite useful.</p>\n<p>There are also some codes I didn't get time to try out but I am sure they also helped a lot of people, like CXXNET, nnForge, Dr. Graham's sparseConvnet code&nbsp;&nbsp;etc.&nbsp;</p>",
  "messages": [
    {
      "id": 66778,
      "postDate": "2015-03-17T18:46:10.047Z",
      "content": "<p>I am sure most people consider working on this kaggle project as a opportunity to learn&nbsp;to use deep learning. Open source codes is a great way for us to get started.</p>\n<p>For me, yr's (ChenglongChen) caffe code and mike1886's lasagne/nolearn codes which implemented realtime augmentation really helped me a lot.</p>\n<p>I also find Sander's Galaxy Zoo code, nagadomi's cifar10 code quite useful.</p>\n<p>There are also some codes I didn't get time to try out but I am sure they also helped a lot of people, like CXXNET, nnForge, Dr. Graham's sparseConvnet code&nbsp;&nbsp;etc.&nbsp;</p>",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "66778": ""
  }
}