{
  "id": 578081,
  "title": "Problem with submission file",
  "url": "/competitions/beyond-visible-spectrum-ai-for-agriculture-2025/discussion/578081",
  "author_name": "",
  "post_date": "2025-05-08T16:37:53.893115100Z",
  "votes": 1,
  "comment_count": 6,
  "views": 0,
  "content": "<p>Hi, I am trying to upload the submission .csv file, and I've got this error:</p>\n<p>ID column ID not found in submission</p>\n<p>Which I do not understand as the file format is like:</p>\n<p>id,label<br>\nsample1957.npy,51</p>\n<p>Am I doing something wrong?</p>",
  "messages": [
    {
      "id": "3197823",
      "postDate": "05/08/2025 16:37:53",
      "content": "<p>Hi, I am trying to upload the submission .csv file, and I've got this error:</p>\n<p>ID column ID not found in submission</p>\n<p>Which I do not understand as the file format is like:</p>\n<p>id,label<br>\nsample1957.npy,51</p>\n<p>Am I doing something wrong?</p>",
      "rawMarkdown": "Hi, I am trying to upload the submission .csv file, and I've got this error:\n\nID column ID not found in submission\n\nWhich I do not understand as the file format is like:\n\nid,label\nsample1957.npy,51\n\nAm I doing something wrong?",
      "votes": null
    },
    {
      "id": "3197829",
      "postDate": "05/08/2025 16:50:23",
      "content": "<p>youve fixed your error im guessing as youve scored </p>",
      "rawMarkdown": "youve fixed your error im guessing as youve scored",
      "votes": null
    },
    {
      "id": "3200019",
      "postDate": "05/11/2025 21:35:46",
      "content": "<p>Yep, you just need to capitalize id -&gt; ID</p>",
      "rawMarkdown": "Yep, you just need to capitalize id -> ID",
      "votes": null
    },
    {
      "id": "3203190",
      "postDate": "05/16/2025 12:08:06",
      "content": "<p>Yes, I did!.😀</p>",
      "rawMarkdown": "Yes, I did!.😀",
      "votes": null
    },
    {
      "id": "3203191",
      "postDate": "05/16/2025 12:08:34",
      "content": "<p>Yes, I did not capitalize it.</p>",
      "rawMarkdown": "Yes, I did not capitalize it.",
      "votes": null
    },
    {
      "id": "3203885",
      "postDate": "05/17/2025 12:37:43",
      "content": "<p>I'm genuinely impressed by how quickly you climbed to the top of the leaderboard! I've been experimenting with several strategies, but my score has been stuck around 845. Would you mind sharing some insights into what you've done differently? I'd really appreciate any tips or guidance.</p>",
      "rawMarkdown": "I'm genuinely impressed by how quickly you climbed to the top of the leaderboard! I've been experimenting with several strategies, but my score has been stuck around 845. Would you mind sharing some insights into what you've done differently? I'd really appreciate any tips or guidance.",
      "votes": null
    },
    {
      "id": "3204556",
      "postDate": "05/18/2025 13:33:29",
      "content": "<p>Honestly, 845 is a good score. Note that it’s calculated using only 37% of the test data. You can still win. The difference between 845 and 822 in MSE terms isn’t that big for this particular challenge. The competition is still wide open. I’ll tell you what I did, but don’t take it as a tip, as I’m not a super expert. I tried some models that resulted in similar scores, meaning that the model architecture may not be a critical factor. I think it is important to select a model whose size is more or less aligned with the size of the training dataset (you can reduce training time, while achieving similar scores). Strong augmentation and a large number of epochs are helping me a lot. My highest score was achieved with a model trained as a regression problem. I’m now exploring the assumption of treating this challenge as a classification problem. But I fear I am running out of time, as the deadline is approaching 😞.</p>",
      "rawMarkdown": "Honestly, 845 is a good score. Note that it’s calculated using only 37% of the test data. You can still win. The difference between 845 and 822 in MSE terms isn’t that big for this particular challenge. The competition is still wide open. I’ll tell you what I did, but don’t take it as a tip, as I’m not a super expert. I tried some models that resulted in similar scores, meaning that the model architecture may not be a critical factor. I think it is important to select a model whose size is more or less aligned with the size of the training dataset (you can reduce training time, while achieving similar scores). Strong augmentation and a large number of epochs are helping me a lot. My highest score was achieved with a model trained as a regression problem. I’m now exploring the assumption of treating this challenge as a classification problem. But I fear I am running out of time, as the deadline is approaching 😞.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3197829,
      "author_name": "loricbobon",
      "author_url": "",
      "post_date": "05/08/2025 16:50:23",
      "content": "<p>youve fixed your error im guessing as youve scored </p>",
      "votes": null,
      "replies": [
        {
          "id": 3203190,
          "author_name": "sergiosanzrodriguez",
          "author_url": "",
          "post_date": "05/16/2025 12:08:06",
          "content": "<p>Yes, I did!.😀</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 3200019,
      "author_name": "dalloliogm",
      "author_url": "",
      "post_date": "05/11/2025 21:35:46",
      "content": "<p>Yep, you just need to capitalize id -&gt; ID</p>",
      "votes": null,
      "replies": [
        {
          "id": 3203191,
          "author_name": "sergiosanzrodriguez",
          "author_url": "",
          "post_date": "05/16/2025 12:08:34",
          "content": "<p>Yes, I did not capitalize it.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 3203885,
      "author_name": "sarthak24910",
      "author_url": "",
      "post_date": "05/17/2025 12:37:43",
      "content": "<p>I'm genuinely impressed by how quickly you climbed to the top of the leaderboard! I've been experimenting with several strategies, but my score has been stuck around 845. Would you mind sharing some insights into what you've done differently? I'd really appreciate any tips or guidance.</p>",
      "votes": null,
      "replies": [
        {
          "id": 3204556,
          "author_name": "sergiosanzrodriguez",
          "author_url": "",
          "post_date": "05/18/2025 13:33:29",
          "content": "<p>Honestly, 845 is a good score. Note that it’s calculated using only 37% of the test data. You can still win. The difference between 845 and 822 in MSE terms isn’t that big for this particular challenge. The competition is still wide open. I’ll tell you what I did, but don’t take it as a tip, as I’m not a super expert. I tried some models that resulted in similar scores, meaning that the model architecture may not be a critical factor. I think it is important to select a model whose size is more or less aligned with the size of the training dataset (you can reduce training time, while achieving similar scores). Strong augmentation and a large number of epochs are helping me a lot. My highest score was achieved with a model trained as a regression problem. I’m now exploring the assumption of treating this challenge as a classification problem. But I fear I am running out of time, as the deadline is approaching 😞.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "3197823": "Hi, I am trying to upload the submission .csv file, and I've got this error:\n\nID column ID not found in submission\n\nWhich I do not understand as the file format is like:\n\nid,label\nsample1957.npy,51\n\nAm I doing something wrong?",
    "3197829": "youve fixed your error im guessing as youve scored",
    "3200019": "Yep, you just need to capitalize id -> ID",
    "3203190": "Yes, I did!.😀",
    "3203191": "Yes, I did not capitalize it.",
    "3203885": "I'm genuinely impressed by how quickly you climbed to the top of the leaderboard! I've been experimenting with several strategies, but my score has been stuck around 845. Would you mind sharing some insights into what you've done differently? I'd really appreciate any tips or guidance.",
    "3204556": "Honestly, 845 is a good score. Note that it’s calculated using only 37% of the test data. You can still win. The difference between 845 and 822 in MSE terms isn’t that big for this particular challenge. The competition is still wide open. I’ll tell you what I did, but don’t take it as a tip, as I’m not a super expert. I tried some models that resulted in similar scores, meaning that the model architecture may not be a critical factor. I think it is important to select a model whose size is more or less aligned with the size of the training dataset (you can reduce training time, while achieving similar scores). Strong augmentation and a large number of epochs are helping me a lot. My highest score was achieved with a model trained as a regression problem. I’m now exploring the assumption of treating this challenge as a classification problem. But I fear I am running out of time, as the deadline is approaching 😞."
  },
  "source": "meta"
}