{
  "id": 232878,
  "title": "Private Test Set",
  "url": "/competitions/bms-molecular-translation/discussion/232878",
  "author_name": "",
  "post_date": "2021-04-15T20:45:32.025890400Z",
  "votes": 3,
  "comment_count": 2,
  "views": 0,
  "content": "<p><strong>Is there a private test set for this competition?</strong>  Or is the \"private\" test set just a subset of the images in the \"test\" folder?  I don't see this in the documentation.  </p>\n<p>I ask because I wonder if taking the approach of converting all the images to tfrecord will fail on the private leaderboard (since we don't have access to those images)…🤔</p>\n<blockquote>\n  <p>The public leaderboard will be based on the public test set and the private leaderboard will be based on the private test set.</p>\n</blockquote>\n<p><a href=\"https://www.kaggle.com/c/bms-molecular-translation/rules\" target=\"_blank\">https://www.kaggle.com/c/bms-molecular-translation/rules</a></p>\n<p>Sounds like the tfrecord approach could result in private test set issues?</p>",
  "messages": [
    {
      "id": "1275033",
      "postDate": "04/15/2021 20:45:32",
      "content": "<p><strong>Is there a private test set for this competition?</strong>  Or is the \"private\" test set just a subset of the images in the \"test\" folder?  I don't see this in the documentation.  </p>\n<p>I ask because I wonder if taking the approach of converting all the images to tfrecord will fail on the private leaderboard (since we don't have access to those images)…🤔</p>\n<blockquote>\n  <p>The public leaderboard will be based on the public test set and the private leaderboard will be based on the private test set.</p>\n</blockquote>\n<p><a href=\"https://www.kaggle.com/c/bms-molecular-translation/rules\" target=\"_blank\">https://www.kaggle.com/c/bms-molecular-translation/rules</a></p>\n<p>Sounds like the tfrecord approach could result in private test set issues?</p>",
      "rawMarkdown": "**Is there a private test set for this competition?**  Or is the \"private\" test set just a subset of the images in the \"test\" folder?  I don't see this in the documentation.  \n\nI ask because I wonder if taking the approach of converting all the images to tfrecord will fail on the private leaderboard (since we don't have access to those images)...🤔\n\n> The public leaderboard will be based on the public test set and the private leaderboard will be based on the private test set.\n\nhttps://www.kaggle.com/c/bms-molecular-translation/rules\n\nSounds like the tfrecord approach could result in private test set issues?",
      "votes": null
    },
    {
      "id": "1275063",
      "postDate": "04/15/2021 21:43:36",
      "content": "<p>From the leaderboard:</p>\n<pre><code>This leaderboard is calculated with approximately 25% of the test data.\nThe final results will be based on the other 75%, so the final standings may be different.\n</code></pre>\n<p>So the private test set is 75% of the images provided in the test folder.</p>",
      "rawMarkdown": "From the leaderboard:\n\n```\nThis leaderboard is calculated with approximately 25% of the test data.\nThe final results will be based on the other 75%, so the final standings may be different.\n```\n\nSo the private test set is 75% of the images provided in the test folder.",
      "votes": null
    },
    {
      "id": "1275071",
      "postDate": "04/15/2021 22:03:13",
      "content": "<p><a href=\"https://www.kaggle.com/sorkun\" target=\"_blank\">@sorkun</a> Thank you for the quick reply! 👍💯</p>",
      "rawMarkdown": "sorkun Thank you for the quick reply! 👍💯",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1275063,
      "author_name": "sorkun",
      "author_url": "",
      "post_date": "04/15/2021 21:43:36",
      "content": "<p>From the leaderboard:</p>\n<pre><code>This leaderboard is calculated with approximately 25% of the test data.\nThe final results will be based on the other 75%, so the final standings may be different.\n</code></pre>\n<p>So the private test set is 75% of the images provided in the test folder.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1275071,
          "author_name": "yeayates21",
          "author_url": "",
          "post_date": "04/15/2021 22:03:13",
          "content": "<p><a href=\"https://www.kaggle.com/sorkun\" target=\"_blank\">@sorkun</a> Thank you for the quick reply! 👍💯</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1275033": "**Is there a private test set for this competition?**  Or is the \"private\" test set just a subset of the images in the \"test\" folder?  I don't see this in the documentation.  \n\nI ask because I wonder if taking the approach of converting all the images to tfrecord will fail on the private leaderboard (since we don't have access to those images)...🤔\n\n> The public leaderboard will be based on the public test set and the private leaderboard will be based on the private test set.\n\nhttps://www.kaggle.com/c/bms-molecular-translation/rules\n\nSounds like the tfrecord approach could result in private test set issues?",
    "1275063": "From the leaderboard:\n\n```\nThis leaderboard is calculated with approximately 25% of the test data.\nThe final results will be based on the other 75%, so the final standings may be different.\n```\n\nSo the private test set is 75% of the images provided in the test folder.",
    "1275071": "sorkun Thank you for the quick reply! 👍💯"
  },
  "source": "meta"
}