{
  "id": 73308,
  "title": "The Poor Man's Quick, Draw!",
  "url": "/competitions/quickdraw-doodle-recognition/discussion/73308",
  "author_name": "garzillo",
  "post_date": "2018-12-01T19:50:33.132000",
  "votes": 18,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi all,</p>\n\n<p>Given the dataset, I thought it would be fun and some good experience to implement a poor man's version of Google's Quick, Draw! application, which I've just called <a href=\"http://www.makepit.com/doodle\">Doodle!</a>.</p>\n\n<p>It's based on my (measly) .906 LB model, but I think it's nice to see just how good all the top models are performing by actually interacting with one.</p>",
  "messages": [
    {
      "id": 431193,
      "postDate": "2018-12-01T19:50:33.133Z",
      "content": "<p>Hi all,</p>\n\n<p>Given the dataset, I thought it would be fun and some good experience to implement a poor man's version of Google's Quick, Draw! application, which I've just called <a href=\"http://www.makepit.com/doodle\">Doodle!</a>.</p>\n\n<p>It's based on my (measly) .906 LB model, but I think it's nice to see just how good all the top models are performing by actually interacting with one.</p>",
      "rawMarkdown": "Hi all,\n\nGiven the dataset, I thought it would be fun and some good experience to implement a poor man's version of Google's Quick, Draw! application, which I've just called [Doodle!][1].\n\nIt's based on my (measly) .906 LB model, but I think it's nice to see just how good all the top models are performing by actually interacting with one.\n\n  [1]: http://www.makepit.com/doodle",
      "votes": 18
    },
    {
      "id": 431224,
      "postDate": "2018-12-01T21:38:40.330Z",
      "content": "<p>Awesome!</p>",
      "rawMarkdown": "Awesome!",
      "votes": 1,
      "replies": [
        {
          "id": 580530,
          "postDate": "2019-07-20T10:01:19.357Z",
          "content": "<p>Yeach/\n! </p>",
          "rawMarkdown": "Yeach/\n! "
        }
      ]
    }
  ],
  "comments": [
    {
      "id": 431224,
      "author_name": "Lukasz Grad",
      "author_url": "",
      "post_date": "2018-12-01T21:38:40.330000",
      "content": "<p>Awesome!</p>",
      "votes": 1,
      "replies": [
        {
          "id": 580530,
          "author_name": "Pavel Solovev",
          "author_url": "",
          "post_date": "2019-07-20T10:01:19.357000",
          "content": "<p>Yeach/\n! </p>",
          "votes": 0,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "431193": "Hi all,\n\nGiven the dataset, I thought it would be fun and some good experience to implement a poor man's version of Google's Quick, Draw! application, which I've just called [Doodle!][1].\n\nIt's based on my (measly) .906 LB model, but I think it's nice to see just how good all the top models are performing by actually interacting with one.\n\n  [1]: http://www.makepit.com/doodle",
    "431224": "Awesome!"
  }
}