{
  "id": 130819,
  "title": "use 3d printer to collect train data ....",
  "url": "/competitions/bengaliai-cv19/discussion/130819",
  "author_name": "",
  "post_date": "2020-02-16T15:00:38.587668Z",
  "votes": 13,
  "comment_count": 5,
  "views": 0,
  "content": "<p><a href=\"https://www.youtube.com/watch?v=EtLYxHRVkRc\">https://www.youtube.com/watch?v=EtLYxHRVkRc</a></p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F113660%2F1c6e867777317b665b355fddc0554a29%2FSelection_078.png?generation=1581865493794900&amp;alt=media\" alt=\"\">\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F113660%2Fd812d2e1b5f8423330a54aa2b166e3ea%2FSelection_079.png?generation=1581865495486700&amp;alt=media\" alt=\"\"></p>",
  "messages": [
    {
      "id": "747542",
      "postDate": "02/16/2020 15:00:38",
      "content": "<p><a href=\"https://www.youtube.com/watch?v=EtLYxHRVkRc\">https://www.youtube.com/watch?v=EtLYxHRVkRc</a></p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F113660%2F1c6e867777317b665b355fddc0554a29%2FSelection_078.png?generation=1581865493794900&amp;alt=media\" alt=\"\">\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F113660%2Fd812d2e1b5f8423330a54aa2b166e3ea%2FSelection_079.png?generation=1581865495486700&amp;alt=media\" alt=\"\"></p>",
      "rawMarkdown": "https://www.youtube.com/watch?v=EtLYxHRVkRc\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F113660%2F1c6e867777317b665b355fddc0554a29%2FSelection_078.png?generation=1581865493794900&amp;alt=media)\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F113660%2Fd812d2e1b5f8423330a54aa2b166e3ea%2FSelection_079.png?generation=1581865495486700&amp;alt=media)",
      "votes": null
    },
    {
      "id": "747562",
      "postDate": "02/16/2020 15:10:40",
      "content": "<p>the parameter space of character\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F113660%2Fcecc94f8250415bb5bce45d7a95d2cef%2FSelection_080.png?generation=1581865838114263&amp;alt=media\" alt=\"\"></p>",
      "rawMarkdown": "the parameter space of character\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F113660%2Fcecc94f8250415bb5bce45d7a95d2cef%2FSelection_080.png?generation=1581865838114263&amp;alt=media)",
      "votes": null
    },
    {
      "id": "747564",
      "postDate": "02/16/2020 15:12:33",
      "content": "<p>a deep learning method to convert input image to the stroke parameters would be interesting</p>",
      "rawMarkdown": "a deep learning method to convert input image to the stroke parameters would be interesting",
      "votes": null
    },
    {
      "id": "751476",
      "postDate": "02/20/2020 08:14:37",
      "content": "<p>Wow</p>",
      "rawMarkdown": "Wow",
      "votes": null
    },
    {
      "id": "751655",
      "postDate": "02/20/2020 11:20:22",
      "content": "<p>bahut hard</p>",
      "rawMarkdown": "bahut hard",
      "votes": null
    },
    {
      "id": "755168",
      "postDate": "02/24/2020 14:11:15",
      "content": "<p>Hi <a href=\"/hengck23\">@hengck23</a> Are you still optimize 68x118 model? I had tried many methods such as change tails augment focal loss and so on, but it seems hard to beyond 0.9795(my best lb score). 😣  </p>",
      "rawMarkdown": "Hi @hengck23 Are you still optimize 68x118 model? I had tried many methods such as change tails augment focal loss and so on, but it seems hard to beyond 0.9795(my best lb score). 😣",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 747562,
      "author_name": "hengck23",
      "author_url": "",
      "post_date": "02/16/2020 15:10:40",
      "content": "<p>the parameter space of character\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F113660%2Fcecc94f8250415bb5bce45d7a95d2cef%2FSelection_080.png?generation=1581865838114263&amp;alt=media\" alt=\"\"></p>",
      "votes": null,
      "replies": [
        {
          "id": 747564,
          "author_name": "hengck23",
          "author_url": "",
          "post_date": "02/16/2020 15:12:33",
          "content": "<p>a deep learning method to convert input image to the stroke parameters would be interesting</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 755168,
          "author_name": "cswwp347724",
          "author_url": "",
          "post_date": "02/24/2020 14:11:15",
          "content": "<p>Hi <a href=\"/hengck23\">@hengck23</a> Are you still optimize 68x118 model? I had tried many methods such as change tails augment focal loss and so on, but it seems hard to beyond 0.9795(my best lb score). 😣  </p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 751476,
      "author_name": "",
      "author_url": "",
      "post_date": "02/20/2020 08:14:37",
      "content": "<p>Wow</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 751655,
      "author_name": "tusharsethi1988",
      "author_url": "",
      "post_date": "02/20/2020 11:20:22",
      "content": "<p>bahut hard</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "747542": "https://www.youtube.com/watch?v=EtLYxHRVkRc\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F113660%2F1c6e867777317b665b355fddc0554a29%2FSelection_078.png?generation=1581865493794900&amp;alt=media)\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F113660%2Fd812d2e1b5f8423330a54aa2b166e3ea%2FSelection_079.png?generation=1581865495486700&amp;alt=media)",
    "747562": "the parameter space of character\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F113660%2Fcecc94f8250415bb5bce45d7a95d2cef%2FSelection_080.png?generation=1581865838114263&amp;alt=media)",
    "747564": "a deep learning method to convert input image to the stroke parameters would be interesting",
    "751476": "Wow",
    "751655": "bahut hard",
    "755168": "Hi @hengck23 Are you still optimize 68x118 model? I had tried many methods such as change tails augment focal loss and so on, but it seems hard to beyond 0.9795(my best lb score). 😣"
  },
  "source": "meta"
}