{
  "id": 507568,
  "title": "How does Notebook Submission works at backend ?",
  "url": "/competitions/home-credit-credit-risk-model-stability/discussion/507568",
  "author_name": "",
  "post_date": "2024-05-26T11:29:34.413087500Z",
  "votes": 1,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Accounting to me , which reading test data with code in the same location kaggle will replace with new public / private test files &amp; submission will have new public / private test data loaded which are 90 % size of train data !!!  Is my above assumption is correct ? please share your related thoughts or facts !!</p>",
  "messages": [
    {
      "id": "2837231",
      "postDate": "05/26/2024 11:29:34",
      "content": "<p>Accounting to me , which reading test data with code in the same location kaggle will replace with new public / private test files &amp; submission will have new public / private test data loaded which are 90 % size of train data !!!  Is my above assumption is correct ? please share your related thoughts or facts !!</p>",
      "rawMarkdown": "Accounting to me , which reading test data with code in the same location kaggle will replace with new public / private test files & submission will have new public / private test data loaded which are 90 % size of train data !!!  Is my above assumption is correct ? please share your related thoughts or facts !!",
      "votes": null
    },
    {
      "id": "2837232",
      "postDate": "05/26/2024 11:34:06",
      "content": "<p>Yes you are right. When one submits to the leaderboard, one's kernel runs on the sample test set that we can see and then immediately runs separately on the <strong>actual test set</strong>. The public and private scores are created and the public score is displayed. <br>\nWhen the competition will end 2 days later, we will be made aware of the private score also <a href=\"https://www.kaggle.com/kunduruanil\" target=\"_blank\">@kunduruanil</a> </p>",
      "rawMarkdown": "Yes you are right. When one submits to the leaderboard, one's kernel runs on the sample test set that we can see and then immediately runs separately on the **actual test set**. The public and private scores are created and the public score is displayed. \nWhen the competition will end 2 days later, we will be made aware of the private score also @kunduruanil",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2837232,
      "author_name": "ravi20076",
      "author_url": "",
      "post_date": "05/26/2024 11:34:06",
      "content": "<p>Yes you are right. When one submits to the leaderboard, one's kernel runs on the sample test set that we can see and then immediately runs separately on the <strong>actual test set</strong>. The public and private scores are created and the public score is displayed. <br>\nWhen the competition will end 2 days later, we will be made aware of the private score also <a href=\"https://www.kaggle.com/kunduruanil\" target=\"_blank\">@kunduruanil</a> </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2837231": "Accounting to me , which reading test data with code in the same location kaggle will replace with new public / private test files & submission will have new public / private test data loaded which are 90 % size of train data !!!  Is my above assumption is correct ? please share your related thoughts or facts !!",
    "2837232": "Yes you are right. When one submits to the leaderboard, one's kernel runs on the sample test set that we can see and then immediately runs separately on the **actual test set**. The public and private scores are created and the public score is displayed. \nWhen the competition will end 2 days later, we will be made aware of the private score also @kunduruanil"
  },
  "source": "meta"
}