{
  "id": 354595,
  "title": "A Stable Competition?",
  "url": "/competitions/hubmap-organ-segmentation/discussion/354595",
  "author_name": "",
  "post_date": "2022-09-23T02:47:13.581581300Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I heard somebody said they got Public LB around 0.8 and gives Private LB around 0.79;</p>\n<p>For me, I didn't put enough time on this match, but all tested single model with about 0.04 Public-Private LB difference.<br>\nFrom the leaderboard, I do not find a huge shake.<br>\nI found the most advanced notebook is: <a href=\"https://www.kaggle.com/code/jinbei03050305/fork-of-inference-hubmap-fast-ai-starter-74e8ea/notebook?scriptVersionId=105102452\" target=\"_blank\">https://www.kaggle.com/code/jinbei03050305/fork-of-inference-hubmap-fast-ai-starter-74e8ea/notebook?scriptVersionId=105102452</a></p>\n<p>anyone have different idea? Please leave your comment :)</p>",
  "messages": [
    {
      "id": "1951346",
      "postDate": "09/23/2022 02:47:13",
      "content": "<p>I heard somebody said they got Public LB around 0.8 and gives Private LB around 0.79;</p>\n<p>For me, I didn't put enough time on this match, but all tested single model with about 0.04 Public-Private LB difference.<br>\nFrom the leaderboard, I do not find a huge shake.<br>\nI found the most advanced notebook is: <a href=\"https://www.kaggle.com/code/jinbei03050305/fork-of-inference-hubmap-fast-ai-starter-74e8ea/notebook?scriptVersionId=105102452\" target=\"_blank\">https://www.kaggle.com/code/jinbei03050305/fork-of-inference-hubmap-fast-ai-starter-74e8ea/notebook?scriptVersionId=105102452</a></p>\n<p>anyone have different idea? Please leave your comment :)</p>",
      "rawMarkdown": "I heard somebody said they got Public LB around 0.8 and gives Private LB around 0.79;\n\nFor me, I didn't put enough time on this match, but all tested single model with about 0.04 Public-Private LB difference.\nFrom the leaderboard, I do not find a huge shake.\nI found the most advanced notebook is: https://www.kaggle.com/code/jinbei03050305/fork-of-inference-hubmap-fast-ai-starter-74e8ea/notebook?scriptVersionId=105102452\n\nanyone have different idea? Please leave your comment :)",
      "votes": null
    },
    {
      "id": "1951357",
      "postDate": "09/23/2022 02:53:03",
      "content": "<p>I think it's only because the author got his notebook scored with \"Hubmap\" dataset. So, I'd rather say he played smart (easy to check his score with unseen decimal score with his rank position)</p>",
      "rawMarkdown": "I think it's only because the author got his notebook scored with \"Hubmap\" dataset. So, I'd rather say he played smart (easy to check his score with unseen decimal score with his rank position)",
      "votes": null
    },
    {
      "id": "1951371",
      "postDate": "09/23/2022 03:00:48",
      "content": "<p>probably, but he didn't use Public LB for checking his score. He must trust his CV score very much.</p>",
      "rawMarkdown": "probably, but he didn't use Public LB for checking his score. He must trust his CV score very much.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1951357,
      "author_name": "cheulkay",
      "author_url": "",
      "post_date": "09/23/2022 02:53:03",
      "content": "<p>I think it's only because the author got his notebook scored with \"Hubmap\" dataset. So, I'd rather say he played smart (easy to check his score with unseen decimal score with his rank position)</p>",
      "votes": null,
      "replies": [
        {
          "id": 1951371,
          "author_name": "dwchen",
          "author_url": "",
          "post_date": "09/23/2022 03:00:48",
          "content": "<p>probably, but he didn't use Public LB for checking his score. He must trust his CV score very much.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1951346": "I heard somebody said they got Public LB around 0.8 and gives Private LB around 0.79;\n\nFor me, I didn't put enough time on this match, but all tested single model with about 0.04 Public-Private LB difference.\nFrom the leaderboard, I do not find a huge shake.\nI found the most advanced notebook is: https://www.kaggle.com/code/jinbei03050305/fork-of-inference-hubmap-fast-ai-starter-74e8ea/notebook?scriptVersionId=105102452\n\nanyone have different idea? Please leave your comment :)",
    "1951357": "I think it's only because the author got his notebook scored with \"Hubmap\" dataset. So, I'd rather say he played smart (easy to check his score with unseen decimal score with his rank position)",
    "1951371": "probably, but he didn't use Public LB for checking his score. He must trust his CV score very much."
  },
  "source": "meta"
}