{
  "id": 514200,
  "title": "About new test set",
  "url": "/competitions/leap-atmospheric-physics-ai-climsim/discussion/514200",
  "author_name": "Gunes Evitan",
  "post_date": "2024-06-23T07:09:54.782000",
  "votes": 9,
  "comment_count": 6,
  "views": 0,
  "content": "<p>I want to ask how different is the new test in terms of locations and timestamps. I saw some people are reporting widened CV/LB gap which happens to me as well. There is almost nothing changed in my validation score but LB score is decreased by 0.01 for all models consistently.</p>",
  "messages": [
    {
      "id": 2885593,
      "postDate": "2024-06-23T07:09:54.783Z",
      "content": "<p>I want to ask how different is the new test in terms of locations and timestamps. I saw some people are reporting widened CV/LB gap which happens to me as well. There is almost nothing changed in my validation score but LB score is decreased by 0.01 for all models consistently.</p>",
      "rawMarkdown": "I want to ask how different is the new test in terms of locations and timestamps. I saw some people are reporting widened CV/LB gap which happens to me as well. There is almost nothing changed in my validation score but LB score is decreased by 0.01 for all models consistently.",
      "votes": 9
    },
    {
      "id": 2886083,
      "postDate": "2024-06-23T12:47:13.533Z",
      "content": "<p>have you solved the question? I also encountered this problem</p>",
      "rawMarkdown": "have you solved the question? I also encountered this problem",
      "replies": [
        {
          "id": 2887667,
          "postDate": "2024-06-24T10:06:52.997Z",
          "content": "<p>Nope I haven't solved it yet.</p>",
          "rawMarkdown": "Nope I haven't solved it yet."
        }
      ]
    },
    {
      "id": 2885945,
      "postDate": "2024-06-23T11:05:08.743Z",
      "rawMarkdown": "",
      "votes": -1,
      "isDeleted": true,
      "replies": [
        {
          "id": 2885951,
          "postDate": "2024-06-23T11:08:22.333Z",
          "content": "<p>Yes, two more columns have non-zero weights now, but applying the previous and current weights on predictions and targets yields almost identical scores for me. </p>",
          "rawMarkdown": "Yes, two more columns have non-zero weights now, but applying the previous and current weights on predictions and targets yields almost identical scores for me. ",
          "replies": [
            {
              "id": 2885963,
              "postDate": "2024-06-23T11:17:43.603Z",
              "content": "<p>It's three new columns.<br>\nDid you remember to apply trick on said three columns?</p>",
              "rawMarkdown": "It's three new columns.\nDid you remember to apply trick on said three columns?"
            },
            {
              "id": 2885990,
              "postDate": "2024-06-23T11:40:09.147Z",
              "content": "<p>Yes, I always overwrite 27 targets located between index 120 and 146 (both inclusive) if you are asking that.</p>",
              "rawMarkdown": "Yes, I always overwrite 27 targets located between index 120 and 146 (both inclusive) if you are asking that."
            }
          ]
        }
      ]
    }
  ],
  "comments": [
    {
      "id": 2886083,
      "author_name": "Starry",
      "author_url": "",
      "post_date": "2024-06-23T12:47:13.533000",
      "content": "<p>have you solved the question? I also encountered this problem</p>",
      "votes": 0,
      "replies": [
        {
          "id": 2887667,
          "author_name": "Gunes Evitan",
          "author_url": "",
          "post_date": "2024-06-24T10:06:52.997000",
          "content": "<p>Nope I haven't solved it yet.</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 2885945,
      "author_name": "",
      "author_url": "",
      "post_date": "2024-06-23T11:05:08.743000",
      "content": "",
      "votes": -1,
      "replies": [
        {
          "id": 2885951,
          "author_name": "Gunes Evitan",
          "author_url": "",
          "post_date": "2024-06-23T11:08:22.333000",
          "content": "<p>Yes, two more columns have non-zero weights now, but applying the previous and current weights on predictions and targets yields almost identical scores for me. </p>",
          "votes": 0,
          "replies": [
            {
              "id": 2885963,
              "author_name": "greySnow",
              "author_url": "",
              "post_date": "2024-06-23T11:17:43.603000",
              "content": "<p>It's three new columns.<br>\nDid you remember to apply trick on said three columns?</p>",
              "votes": 0,
              "replies": []
            },
            {
              "id": 2885990,
              "author_name": "Gunes Evitan",
              "author_url": "",
              "post_date": "2024-06-23T11:40:09.147000",
              "content": "<p>Yes, I always overwrite 27 targets located between index 120 and 146 (both inclusive) if you are asking that.</p>",
              "votes": 0,
              "replies": []
            }
          ]
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2885593": "I want to ask how different is the new test in terms of locations and timestamps. I saw some people are reporting widened CV/LB gap which happens to me as well. There is almost nothing changed in my validation score but LB score is decreased by 0.01 for all models consistently.",
    "2886083": "have you solved the question? I also encountered this problem",
    "2885945": ""
  }
}