{
  "id": 683786,
  "title": "Question about runtimes during private runs",
  "url": "/competitions/deep-past-initiative-machine-translation/discussion/683786",
  "author_name": "",
  "post_date": "2026-03-22T08:54:08.004936100Z",
  "votes": 1,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Hi <a href=\"https://www.kaggle.com/deeppast\" target=\"_blank\">@deeppast</a> </p>\n<p>Just a quick question. The current public set submissions, the time that is being taken, does the test set have 34% of the data? Or is the run done on the entire test set but evals are only calculated on the 34% ?</p>\n<p>I understand that this might be late to bring this up it is really important. If the runs occur only on the 34% of data, teams who have runtimes larger than 3-4 hours would need to spend the last couple days left, optimizing their solution to fit the 9 hour runtime limit. </p>",
  "messages": [
    {
      "id": "3426262",
      "postDate": "03/22/2026 08:54:08",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/deeppast\" target=\"_blank\">@deeppast</a> </p>\n<p>Just a quick question. The current public set submissions, the time that is being taken, does the test set have 34% of the data? Or is the run done on the entire test set but evals are only calculated on the 34% ?</p>\n<p>I understand that this might be late to bring this up it is really important. If the runs occur only on the 34% of data, teams who have runtimes larger than 3-4 hours would need to spend the last couple days left, optimizing their solution to fit the 9 hour runtime limit. </p>",
      "rawMarkdown": "Hi @deeppast \n\nJust a quick question. The current public set submissions, the time that is being taken, does the test set have 34% of the data? Or is the run done on the entire test set but evals are only calculated on the 34% ?\n\nI understand that this might be late to bring this up it is really important. If the runs occur only on the 34% of data, teams who have runtimes larger than 3-4 hours would need to spend the last couple days left, optimizing their solution to fit the 9 hour runtime limit.",
      "votes": null
    },
    {
      "id": "3426263",
      "postDate": "03/22/2026 08:55:48",
      "content": "<p>The run is done on the entire set, if your submission went through it's already been scored for the final result.</p>",
      "rawMarkdown": "The run is done on the entire set, if your submission went through it's already been scored for the final result.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3426263,
      "author_name": "tennogh",
      "author_url": "",
      "post_date": "03/22/2026 08:55:48",
      "content": "<p>The run is done on the entire set, if your submission went through it's already been scored for the final result.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3426262": "Hi @deeppast \n\nJust a quick question. The current public set submissions, the time that is being taken, does the test set have 34% of the data? Or is the run done on the entire test set but evals are only calculated on the 34% ?\n\nI understand that this might be late to bring this up it is really important. If the runs occur only on the 34% of data, teams who have runtimes larger than 3-4 hours would need to spend the last couple days left, optimizing their solution to fit the 9 hour runtime limit.",
    "3426263": "The run is done on the entire set, if your submission went through it's already been scored for the final result."
  },
  "source": "meta"
}