{
  "id": 167038,
  "title": "Runtime of submissions to be displayed!",
  "url": "/competitions/prostate-cancer-grade-assessment/discussion/167038",
  "author_name": "Michaël Karpe",
  "post_date": "2020-07-15T02:20:47.451000",
  "votes": 1,
  "comment_count": 5,
  "views": 0,
  "content": "<p>As a follow-up to these discussions <a href=\"https://www.kaggle.com/product-feedback/126239\">here (original post)</a> and <a href=\"https://www.kaggle.com/product-feedback/132622\">here (duplicate)</a>, I think such a feature would be particularly useful for code competitions like this one.</p>\n\n<p>Have you ever needed to know how long it would take to complete your submission before?</p>\n\n<p>Feel free to reopen the discussion on the original post!</p>",
  "messages": [
    {
      "id": 931610,
      "postDate": "2020-07-16T09:57:27.513Z",
      "content": "<p>This is not done on purpose as it would give you more room for probing.</p>",
      "rawMarkdown": "This is not done on purpose as it would give you more room for probing.",
      "votes": 3,
      "replies": [
        {
          "id": 935875,
          "postDate": "2020-07-19T18:17:36.943Z",
          "content": "<p>Oh ok, I'm quite new to Kaggle, so I didn't think about that… Thank you for your reply!</p>",
          "rawMarkdown": "Oh ok, I'm quite new to Kaggle, so I didn't think about that... Thank you for your reply!"
        }
      ]
    },
    {
      "id": 929833,
      "postDate": "2020-07-15T02:20:47.450Z",
      "content": "<p>As a follow-up to these discussions <a href=\"https://www.kaggle.com/product-feedback/126239\">here (original post)</a> and <a href=\"https://www.kaggle.com/product-feedback/132622\">here (duplicate)</a>, I think such a feature would be particularly useful for code competitions like this one.</p>\n\n<p>Have you ever needed to know how long it would take to complete your submission before?</p>\n\n<p>Feel free to reopen the discussion on the original post!</p>",
      "rawMarkdown": "As a follow-up to these discussions [here (original post)](https://www.kaggle.com/product-feedback/126239) and [here (duplicate)](https://www.kaggle.com/product-feedback/132622), I think such a feature would be particularly useful for code competitions like this one.\n\nHave you ever needed to know how long it would take to complete your submission before?\n\nFeel free to reopen the discussion on the original post!",
      "votes": 1
    },
    {
      "id": 929866,
      "postDate": "2020-07-15T03:15:29.187Z",
      "content": "<p>By the way, when it's written in the code requirements that the condition \"GPU Notebook &lt;= 6 hours run-time\" must be met, is a submission run against 100% of the test data or only against the 42% used for public leaderboard?</p>\n\n<p>In order words, should we expect our submission to be rerun on a bigger set and adapt our runtime consequently so that it remains under 6 hours for GPU notebooks?</p>\n\n<p>Thank you for your help!</p>",
      "rawMarkdown": "By the way, when it's written in the code requirements that the condition \"GPU Notebook &lt;= 6 hours run-time\" must be met, is a submission run against 100% of the test data or only against the 42% used for public leaderboard?\n\nIn order words, should we expect our submission to be rerun on a bigger set and adapt our runtime consequently so that it remains under 6 hours for GPU notebooks?\n\nThank you for your help!",
      "replies": [
        {
          "id": 931569,
          "postDate": "2020-07-16T09:17:26.820Z",
          "content": "<p>When you do submit on the submission file in your notebook output it should run against the entire  hidden test set but the public LB will show the score for the 42% until the end of the competition. So if it completes in 6hrs GPU you should be OK.  I usually try to check one submission while it is running to get an idea of how long it takes - like nn minutes ago under the notebook name and version until the public score appears. </p>",
          "rawMarkdown": "When you do submit on the submission file in your notebook output it should run against the entire  hidden test set but the public LB will show the score for the 42% until the end of the competition. So if it completes in 6hrs GPU you should be OK.  I usually try to check one submission while it is running to get an idea of how long it takes - like nn minutes ago under the notebook name and version until the public score appears. ",
          "votes": 1
        },
        {
          "id": 935876,
          "postDate": "2020-07-19T18:18:11.663Z",
          "content": "<p>Great, thank you for your confirmation!</p>",
          "rawMarkdown": "Great, thank you for your confirmation!"
        }
      ]
    }
  ],
  "comments": [
    {
      "id": 931610,
      "author_name": "Psi",
      "author_url": "",
      "post_date": "2020-07-16T09:57:27.513000",
      "content": "<p>This is not done on purpose as it would give you more room for probing.</p>",
      "votes": 3,
      "replies": [
        {
          "id": 935875,
          "author_name": "Michaël Karpe",
          "author_url": "",
          "post_date": "2020-07-19T18:17:36.943000",
          "content": "<p>Oh ok, I'm quite new to Kaggle, so I didn't think about that… Thank you for your reply!</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 929866,
      "author_name": "Michaël Karpe",
      "author_url": "",
      "post_date": "2020-07-15T03:15:29.187000",
      "content": "<p>By the way, when it's written in the code requirements that the condition \"GPU Notebook &lt;= 6 hours run-time\" must be met, is a submission run against 100% of the test data or only against the 42% used for public leaderboard?</p>\n\n<p>In order words, should we expect our submission to be rerun on a bigger set and adapt our runtime consequently so that it remains under 6 hours for GPU notebooks?</p>\n\n<p>Thank you for your help!</p>",
      "votes": 0,
      "replies": [
        {
          "id": 931569,
          "author_name": "something4kag",
          "author_url": "",
          "post_date": "2020-07-16T09:17:26.820000",
          "content": "<p>When you do submit on the submission file in your notebook output it should run against the entire  hidden test set but the public LB will show the score for the 42% until the end of the competition. So if it completes in 6hrs GPU you should be OK.  I usually try to check one submission while it is running to get an idea of how long it takes - like nn minutes ago under the notebook name and version until the public score appears. </p>",
          "votes": 1,
          "replies": []
        },
        {
          "id": 935876,
          "author_name": "Michaël Karpe",
          "author_url": "",
          "post_date": "2020-07-19T18:18:11.663000",
          "content": "<p>Great, thank you for your confirmation!</p>",
          "votes": 0,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "931610": "This is not done on purpose as it would give you more room for probing.",
    "929833": "As a follow-up to these discussions [here (original post)](https://www.kaggle.com/product-feedback/126239) and [here (duplicate)](https://www.kaggle.com/product-feedback/132622), I think such a feature would be particularly useful for code competitions like this one.\n\nHave you ever needed to know how long it would take to complete your submission before?\n\nFeel free to reopen the discussion on the original post!",
    "929866": "By the way, when it's written in the code requirements that the condition \"GPU Notebook &lt;= 6 hours run-time\" must be met, is a submission run against 100% of the test data or only against the 42% used for public leaderboard?\n\nIn order words, should we expect our submission to be rerun on a bigger set and adapt our runtime consequently so that it remains under 6 hours for GPU notebooks?\n\nThank you for your help!"
  }
}