{
  "id": 53332,
  "title": "Heng CherKeng is too strong",
  "url": "/competitions/data-science-bowl-2018/discussion/53332",
  "author_name": "",
  "post_date": "2018-03-29T13:46:44.222848800Z",
  "votes": 14,
  "comment_count": 9,
  "views": 0,
  "content": "<p>Getting such great LB score without any sub is really amazing ;)</p>\n\n<p><img src=\"https://storage.googleapis.com/kaggle-forum-message-attachments/305837/8899/Screenshot-2018-3-29%202018%20Data%20Science%20Bowl%20Kaggle.png\" alt=\"no sub\"></p>\n\n<p>For those missing it, the 0 on the right of the image is the number of submissions as reported by Kaggle.  There is clearly a bug there.</p>",
  "messages": [
    {
      "id": "305837",
      "postDate": "03/29/2018 13:46:44",
      "content": "<p>Getting such great LB score without any sub is really amazing ;)</p>\n\n<p><img src=\"https://storage.googleapis.com/kaggle-forum-message-attachments/305837/8899/Screenshot-2018-3-29%202018%20Data%20Science%20Bowl%20Kaggle.png\" alt=\"no sub\"></p>\n\n<p>For those missing it, the 0 on the right of the image is the number of submissions as reported by Kaggle.  There is clearly a bug there.</p>",
      "rawMarkdown": "Getting such great LB score without any sub is really amazing ;)\n\n![no sub][1]\n\nFor those missing it, the 0 on the right of the image is the number of submissions as reported by Kaggle.  There is clearly a bug there.\n\n  [1]: https://kaggle2.blob.core.windows.net/forum-message-attachments/305837/8899/Screenshot-2018-3-29%202018%20Data%20Science%20Bowl%20Kaggle.png",
      "votes": null
    },
    {
      "id": "305841",
      "postDate": "03/29/2018 13:52:38",
      "content": "<p>I ran his code but only got 0.47 LB. I think he has many tips that we can learn after he published it. Thanks Heng for sharing your idea</p>",
      "rawMarkdown": "I ran his code but only got 0.47 LB. I think he has many tips that we can learn after he published it. Thanks Heng for sharing your idea",
      "votes": null
    },
    {
      "id": "305853",
      "postDate": "03/29/2018 14:15:43",
      "content": "<p>I thought that only Chuck Norris can do such things.</p>",
      "rawMarkdown": "I thought that only Chuck Norris can do such things.",
      "votes": null
    },
    {
      "id": "306013",
      "postDate": "03/29/2018 18:32:47",
      "content": "<p>I think his score is &gt; 1.0. Kaggle just didn't prepare for that.</p>",
      "rawMarkdown": "I think his score is &gt; 1.0. Kaggle just didn't prepare for that.",
      "votes": null
    },
    {
      "id": "306040",
      "postDate": "03/29/2018 19:25:38",
      "content": "<p>Maybe it's just because they change their team name, submission count is reset to be 0.</p>",
      "rawMarkdown": "Maybe it's just because they change their team name, submission count is reset to be 0.",
      "votes": null
    },
    {
      "id": "306097",
      "postDate": "03/29/2018 21:09:26",
      "content": "<p>Nope, this is a bug.</p>",
      "rawMarkdown": "Nope, this is a bug.",
      "votes": null
    },
    {
      "id": "306229",
      "postDate": "03/30/2018 04:26:34",
      "content": "<p>I try heng's code for serveral epochs but only got 0.3, the performance is bad, How many epochs did you train to get 0.47 ？ And what about your hyperparameter set?(Please ignore me if this is a secret )</p>",
      "rawMarkdown": "I try heng's code for serveral epochs but only got 0.3, the performance is bad, How many epochs did you train to get 0.47 ？ And what about your hyperparameter set?(Please ignore me if this is a secret )",
      "votes": null
    },
    {
      "id": "306247",
      "postDate": "03/30/2018 05:25:44",
      "content": "<p>This topic is about a bug in Kaggle display...  I suggest you ask questions on Heng's code in the (many) threads he created.</p>",
      "rawMarkdown": "This topic is about a bug in Kaggle display...  I suggest you ask questions on Heng's code in the (many) threads he created.",
      "votes": null
    },
    {
      "id": "306504",
      "postDate": "03/30/2018 14:25:42",
      "content": "<p>@CPMP thanks for screen capturing. It is going to be a memorable moment for me.</p>\n\n<p>What an interesting kaggle bug!</p>",
      "rawMarkdown": "CPMP thanks for screen capturing. It is going to be a memorable moment for me.\n\nWhat an interesting kaggle bug!",
      "votes": null
    },
    {
      "id": "306910",
      "postDate": "03/31/2018 09:42:05",
      "content": "<p>Oh I'm sorry</p>",
      "rawMarkdown": "Oh I'm sorry",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 305841,
      "author_name": "john1231983",
      "author_url": "",
      "post_date": "03/29/2018 13:52:38",
      "content": "<p>I ran his code but only got 0.47 LB. I think he has many tips that we can learn after he published it. Thanks Heng for sharing your idea</p>",
      "votes": null,
      "replies": [
        {
          "id": 306229,
          "author_name": "ermiao",
          "author_url": "",
          "post_date": "03/30/2018 04:26:34",
          "content": "<p>I try heng's code for serveral epochs but only got 0.3, the performance is bad, How many epochs did you train to get 0.47 ？ And what about your hyperparameter set?(Please ignore me if this is a secret )</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 306247,
          "author_name": "cpmpml",
          "author_url": "",
          "post_date": "03/30/2018 05:25:44",
          "content": "<p>This topic is about a bug in Kaggle display...  I suggest you ask questions on Heng's code in the (many) threads he created.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 306910,
          "author_name": "ermiao",
          "author_url": "",
          "post_date": "03/31/2018 09:42:05",
          "content": "<p>Oh I'm sorry</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 305853,
      "author_name": "gpreda",
      "author_url": "",
      "post_date": "03/29/2018 14:15:43",
      "content": "<p>I thought that only Chuck Norris can do such things.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 306013,
      "author_name": "seesee",
      "author_url": "",
      "post_date": "03/29/2018 18:32:47",
      "content": "<p>I think his score is &gt; 1.0. Kaggle just didn't prepare for that.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 306040,
      "author_name": "why94nb",
      "author_url": "",
      "post_date": "03/29/2018 19:25:38",
      "content": "<p>Maybe it's just because they change their team name, submission count is reset to be 0.</p>",
      "votes": null,
      "replies": [
        {
          "id": 306097,
          "author_name": "cpmpml",
          "author_url": "",
          "post_date": "03/29/2018 21:09:26",
          "content": "<p>Nope, this is a bug.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 306504,
      "author_name": "hengck23",
      "author_url": "",
      "post_date": "03/30/2018 14:25:42",
      "content": "<p>@CPMP thanks for screen capturing. It is going to be a memorable moment for me.</p>\n\n<p>What an interesting kaggle bug!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "305837": "Getting such great LB score without any sub is really amazing ;)\n\n![no sub][1]\n\nFor those missing it, the 0 on the right of the image is the number of submissions as reported by Kaggle.  There is clearly a bug there.\n\n  [1]: https://kaggle2.blob.core.windows.net/forum-message-attachments/305837/8899/Screenshot-2018-3-29%202018%20Data%20Science%20Bowl%20Kaggle.png",
    "305841": "I ran his code but only got 0.47 LB. I think he has many tips that we can learn after he published it. Thanks Heng for sharing your idea",
    "305853": "I thought that only Chuck Norris can do such things.",
    "306013": "I think his score is &gt; 1.0. Kaggle just didn't prepare for that.",
    "306040": "Maybe it's just because they change their team name, submission count is reset to be 0.",
    "306097": "Nope, this is a bug.",
    "306229": "I try heng's code for serveral epochs but only got 0.3, the performance is bad, How many epochs did you train to get 0.47 ？ And what about your hyperparameter set?(Please ignore me if this is a secret )",
    "306247": "This topic is about a bug in Kaggle display...  I suggest you ask questions on Heng's code in the (many) threads he created.",
    "306504": "CPMP thanks for screen capturing. It is going to be a memorable moment for me.\n\nWhat an interesting kaggle bug!",
    "306910": "Oh I'm sorry"
  },
  "source": "meta"
}