{
  "id": 78132,
  "title": "WOW! Two 0.712 teams!",
  "url": "/competitions/quora-insincere-questions-classification/discussion/78132",
  "author_name": "Shujian Liu",
  "post_date": "2019-01-20T06:50:29.759000",
  "votes": 8,
  "comment_count": 3,
  "views": 0,
  "content": "",
  "messages": [
    {
      "id": 458647,
      "postDate": "2019-01-20T06:50:29.760Z",
      "rawMarkdown": "",
      "votes": 7
    },
    {
      "id": 458753,
      "postDate": "2019-01-20T12:47:51.853Z",
      "content": "<p>Congrats to these two teams! \nThey have been steadily advancing in past days and I guess they will continue to do so. </p>",
      "rawMarkdown": "Congrats to these two teams! \nThey have been steadily advancing in past days and I guess they will continue to do so. ",
      "votes": 2,
      "isDeleted": true,
      "replies": [
        {
          "id": 458783,
          "postDate": "2019-01-20T13:42:49.440Z",
          "content": "<p>Congrat to you too for being able to pass the .707 barrier! (In my case, I still stuck at the same barrier haha)</p>",
          "rawMarkdown": "Congrat to you too for being able to pass the .707 barrier! (In my case, I still stuck at the same barrier haha)",
          "votes": 1
        },
        {
          "id": 458824,
          "postDate": "2019-01-20T14:52:06.700Z",
          "content": "<p>Thanks; yet I still have a hard time to replicate this score due to small differences in predicted probs across runs. \nWhat we need is <strong>keras.put_seeds_absolutely_everywhere()</strong> but it is not yet implemented :-)</p>",
          "rawMarkdown": "Thanks; yet I still have a hard time to replicate this score due to small differences in predicted probs across runs. \nWhat we need is **keras.put_seeds_absolutely_everywhere()** but it is not yet implemented :-)\n\n\n",
          "votes": 8,
          "isDeleted": true
        }
      ]
    }
  ],
  "comments": [
    {
      "id": 458753,
      "author_name": "",
      "author_url": "",
      "post_date": "2019-01-20T12:47:51.853000",
      "content": "<p>Congrats to these two teams! \nThey have been steadily advancing in past days and I guess they will continue to do so. </p>",
      "votes": 2,
      "replies": [
        {
          "id": 458783,
          "author_name": "Neuron Engineer",
          "author_url": "",
          "post_date": "2019-01-20T13:42:49.440000",
          "content": "<p>Congrat to you too for being able to pass the .707 barrier! (In my case, I still stuck at the same barrier haha)</p>",
          "votes": 1,
          "replies": []
        },
        {
          "id": 458824,
          "author_name": "",
          "author_url": "",
          "post_date": "2019-01-20T14:52:06.700000",
          "content": "<p>Thanks; yet I still have a hard time to replicate this score due to small differences in predicted probs across runs. \nWhat we need is <strong>keras.put_seeds_absolutely_everywhere()</strong> but it is not yet implemented :-)</p>",
          "votes": 8,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "458647": "",
    "458753": "Congrats to these two teams! \nThey have been steadily advancing in past days and I guess they will continue to do so. "
  }
}