{
  "id": 669030,
  "title": "What is the test set over which computing the prediction?",
  "url": "/competitions/beyond-visible-spectrum-ai-for-agriculture-2026p2/discussion/669030",
  "author_name": "",
  "post_date": "2026-01-20T09:22:01.334152Z",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hello, The dataset structure is totally unclear. Please describe exactly what each folder contains, in particular I understand that the test set over which I compute the final prediction is inside the folder archive/share/val …, but the submission form asks to upload a csv with 40 rows only, but there are 432 record in archive/share/val.\nWould you please specify where the test set is located? </p>\n<p>Thanks.</p>",
  "messages": [
    {
      "id": "3394001",
      "postDate": "01/20/2026 09:22:01",
      "content": "<p>Hello, The dataset structure is totally unclear. Please describe exactly what each folder contains, in particular I understand that the test set over which I compute the final prediction is inside the folder archive/share/val …, but the submission form asks to upload a csv with 40 rows only, but there are 432 record in archive/share/val.\nWould you please specify where the test set is located? </p>\n<p>Thanks.</p>",
      "rawMarkdown": "Hello, The dataset structure is totally unclear. Please describe exactly what each folder contains, in particular I understand that the test set over which I compute the final prediction is inside the folder archive/share/val ..., but the submission form asks to upload a csv with 40 rows only, but there are 432 record in archive/share/val.\nWould you please specify where the test set is located? \n\nThanks.",
      "votes": null
    },
    {
      "id": "3394010",
      "postDate": "01/20/2026 09:37:36",
      "content": "<p>Hi, you don't need archive file at all. Just predict on val folder </p>",
      "rawMarkdown": "Hi, you don't need archive file at all. Just predict on val folder",
      "votes": null
    },
    {
      "id": "3394012",
      "postDate": "01/20/2026 09:41:57",
      "content": "<p>Do you mean the folder ICPR02/kaggle/evaluation?</p>",
      "rawMarkdown": "Do you mean the folder ICPR02/kaggle/evaluation?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3394010,
      "author_name": "robeson",
      "author_url": "",
      "post_date": "01/20/2026 09:37:36",
      "content": "<p>Hi, you don't need archive file at all. Just predict on val folder </p>",
      "votes": null,
      "replies": [
        {
          "id": 3394012,
          "author_name": "came86",
          "author_url": "",
          "post_date": "01/20/2026 09:41:57",
          "content": "<p>Do you mean the folder ICPR02/kaggle/evaluation?</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "3394001": "Hello, The dataset structure is totally unclear. Please describe exactly what each folder contains, in particular I understand that the test set over which I compute the final prediction is inside the folder archive/share/val ..., but the submission form asks to upload a csv with 40 rows only, but there are 432 record in archive/share/val.\nWould you please specify where the test set is located? \n\nThanks.",
    "3394010": "Hi, you don't need archive file at all. Just predict on val folder",
    "3394012": "Do you mean the folder ICPR02/kaggle/evaluation?"
  },
  "source": "meta"
}