{
  "id": 668205,
  "title": "Test set lines and `text_id`",
  "url": "/competitions/deep-past-initiative-machine-translation/discussion/668205",
  "author_name": "",
  "post_date": "2026-01-15T17:13:28.054823900Z",
  "votes": 1,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Hey everyone,</p>\n<p>I have now come to the part where I am considering utilizing line starts and ends and <code>text_id</code> columns that are found in <code>test.csv</code>. </p>\n<p>I guess the question is both to the organizers and other people:</p>\n<ol>\n<li>Is the hidden test set also going to contain line numbers and text IDs consistently, is that something we will regularly get or is it just one example in our dummy test set?</li>\n<li>Has anyone made use of it directly or indirectly. Like, you could provide model's translation for the previous lines to translate the next lines or smth like it. </li>\n</ol>\n<p>Thanks everyone!</p>",
  "messages": [
    {
      "id": "3391793",
      "postDate": "01/15/2026 17:13:28",
      "content": "<p>Hey everyone,</p>\n<p>I have now come to the part where I am considering utilizing line starts and ends and <code>text_id</code> columns that are found in <code>test.csv</code>. </p>\n<p>I guess the question is both to the organizers and other people:</p>\n<ol>\n<li>Is the hidden test set also going to contain line numbers and text IDs consistently, is that something we will regularly get or is it just one example in our dummy test set?</li>\n<li>Has anyone made use of it directly or indirectly. Like, you could provide model's translation for the previous lines to translate the next lines or smth like it. </li>\n</ol>\n<p>Thanks everyone!</p>",
      "rawMarkdown": "Hey everyone,\n\nI have now come to the part where I am considering utilizing line starts and ends and `text_id` columns that are found in `test.csv`. \n\nI guess the question is both to the organizers and other people:\n\n1. Is the hidden test set also going to contain line numbers and text IDs consistently, is that something we will regularly get or is it just one example in our dummy test set?\n2. Has anyone made use of it directly or indirectly. Like, you could provide model's translation for the previous lines to translate the next lines or smth like it. \n\nThanks everyone!",
      "votes": null
    },
    {
      "id": "3392509",
      "postDate": "01/17/2026 04:41:00",
      "content": "<p>The test set contains for each row: Document IDs, a column for the 'Line_start' and one for 'Line_end'. The rows of text strings are segmented into sentences, for the most part.</p>",
      "rawMarkdown": "The test set contains for each row: Document IDs, a column for the 'Line_start' and one for 'Line_end'. The rows of text strings are segmented into sentences, for the most part.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3392509,
      "author_name": "deeppast",
      "author_url": "",
      "post_date": "01/17/2026 04:41:00",
      "content": "<p>The test set contains for each row: Document IDs, a column for the 'Line_start' and one for 'Line_end'. The rows of text strings are segmented into sentences, for the most part.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3391793": "Hey everyone,\n\nI have now come to the part where I am considering utilizing line starts and ends and `text_id` columns that are found in `test.csv`. \n\nI guess the question is both to the organizers and other people:\n\n1. Is the hidden test set also going to contain line numbers and text IDs consistently, is that something we will regularly get or is it just one example in our dummy test set?\n2. Has anyone made use of it directly or indirectly. Like, you could provide model's translation for the previous lines to translate the next lines or smth like it. \n\nThanks everyone!",
    "3392509": "The test set contains for each row: Document IDs, a column for the 'Line_start' and one for 'Line_end'. The rows of text strings are segmented into sentences, for the most part."
  },
  "source": "meta"
}