{
  "id": 388761,
  "title": "Public Leaderboard Percentage",
  "url": "/competitions/ml-olympiad-dialectrecognition/discussion/388761",
  "author_name": "",
  "post_date": "2023-02-19T12:07:00.887603800Z",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Hi, <br>\nThis is an interesting topic. And a lot of experiments can be done here! However with 1% only as a public score! The leaderboard (as feedback/information source) becomes useless quickly&nbsp; ! For example, my latest tests give&nbsp;less leaderboard score but a higher validation score! </p>\n<p>If I suggest reconsidering&nbsp;the 1%, 99% split (if Kaggle allows it).<br>\nAnyway&nbsp;it's a fun and interesting challenge and I will share my findings in the end</p>",
  "messages": [
    {
      "id": "2150580",
      "postDate": "02/19/2023 12:07:00",
      "content": "<p>Hi, <br>\nThis is an interesting topic. And a lot of experiments can be done here! However with 1% only as a public score! The leaderboard (as feedback/information source) becomes useless quickly&nbsp; ! For example, my latest tests give&nbsp;less leaderboard score but a higher validation score! </p>\n<p>If I suggest reconsidering&nbsp;the 1%, 99% split (if Kaggle allows it).<br>\nAnyway&nbsp;it's a fun and interesting challenge and I will share my findings in the end</p>",
      "rawMarkdown": "Hi, \nThis is an interesting topic. And a lot of experiments can be done here! However with 1% only as a public score! The leaderboard (as feedback/information source) becomes useless quickly  ! For example, my latest tests give less leaderboard score but a higher validation score! \n\nIf I suggest reconsidering the 1%, 99% split (if Kaggle allows it).\nAnyway it's a fun and interesting challenge and I will share my findings in the end",
      "votes": null
    },
    {
      "id": "2154799",
      "postDate": "02/22/2023 08:59:33",
      "content": "<p>Hello <a href=\"https://www.kaggle.com/asalhi\" target=\"_blank\">@asalhi</a> ,</p>\n<p>The percentage of public leaderboard has been raised to 50.</p>\n<p>Looking forward to see impressive results from you.</p>\n<p>Good luck! </p>",
      "rawMarkdown": "Hello @asalhi ,\n\nThe percentage of public leaderboard has been raised to 50.\n\nLooking forward to see impressive results from you.\n\nGood luck!",
      "votes": null
    },
    {
      "id": "2154810",
      "postDate": "02/22/2023 09:11:22",
      "content": "<p>Thanks <a href=\"https://www.kaggle.com/ruqiyas\" target=\"_blank\">@ruqiyas</a> :-) </p>\n<p>You may consider resetting the leaderboard since the current values reflect the best of 1% 99% split </p>",
      "rawMarkdown": "Thanks @ruqiyas :-) \n\nYou may consider resetting the leaderboard since the current values reflect the best of 1% 99% split",
      "votes": null
    },
    {
      "id": "2154850",
      "postDate": "02/22/2023 09:32:09",
      "content": "<p>I don't have the option to do that in the host setting. </p>\n<p>Anyway, don't worry about the order, as it is based on both the public and private leaderboard.</p>",
      "rawMarkdown": "I don't have the option to do that in the host setting. \n\nAnyway, don't worry about the order, as it is based on both the public and private leaderboard.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2154799,
      "author_name": "ruqiyas",
      "author_url": "",
      "post_date": "02/22/2023 08:59:33",
      "content": "<p>Hello <a href=\"https://www.kaggle.com/asalhi\" target=\"_blank\">@asalhi</a> ,</p>\n<p>The percentage of public leaderboard has been raised to 50.</p>\n<p>Looking forward to see impressive results from you.</p>\n<p>Good luck! </p>",
      "votes": null,
      "replies": [
        {
          "id": 2154810,
          "author_name": "asalhi",
          "author_url": "",
          "post_date": "02/22/2023 09:11:22",
          "content": "<p>Thanks <a href=\"https://www.kaggle.com/ruqiyas\" target=\"_blank\">@ruqiyas</a> :-) </p>\n<p>You may consider resetting the leaderboard since the current values reflect the best of 1% 99% split </p>",
          "votes": null,
          "replies": [
            {
              "id": 2154850,
              "author_name": "ruqiyas",
              "author_url": "",
              "post_date": "02/22/2023 09:32:09",
              "content": "<p>I don't have the option to do that in the host setting. </p>\n<p>Anyway, don't worry about the order, as it is based on both the public and private leaderboard.</p>",
              "votes": null,
              "replies": []
            }
          ]
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2150580": "Hi, \nThis is an interesting topic. And a lot of experiments can be done here! However with 1% only as a public score! The leaderboard (as feedback/information source) becomes useless quickly  ! For example, my latest tests give less leaderboard score but a higher validation score! \n\nIf I suggest reconsidering the 1%, 99% split (if Kaggle allows it).\nAnyway it's a fun and interesting challenge and I will share my findings in the end",
    "2154799": "Hello @asalhi ,\n\nThe percentage of public leaderboard has been raised to 50.\n\nLooking forward to see impressive results from you.\n\nGood luck!",
    "2154810": "Thanks @ruqiyas :-) \n\nYou may consider resetting the leaderboard since the current values reflect the best of 1% 99% split",
    "2154850": "I don't have the option to do that in the host setting. \n\nAnyway, don't worry about the order, as it is based on both the public and private leaderboard."
  },
  "source": "meta"
}