{
  "id": 663517,
  "title": "Submission scoring error",
  "url": "/competitions/recodai-luc-scientific-image-forgery-detection/discussion/663517",
  "author_name": "",
  "post_date": "2025-12-18T13:16:27.662463700Z",
  "votes": null,
  "comment_count": 5,
  "views": 0,
  "content": "<p>I keep getting this error on submission:\nYour notebook generated a submission file with incorrect format. Some examples causing this are: wrong number of rows or columns, empty values, an incorrect data type for a value, or invalid submission values from what is expected. See more debugging tips.</p>\n<p>As far as I checked for the submission.csv it properly generates a submission in <code>/kaggle/working/submission.csv</code></p>\n<p>And it contains the correctly formatted content for the given (one) test image in test_images directory:</p>\n<pre><code>Submission shape: (1, 2)\nColumns: ['case_id', 'annotation']\n</code></pre>\n<p>Am I missing anything here? It keeps getting rejected with Submission scoring error. Would appreciate suggestions for this fix.</p>",
  "messages": [
    {
      "id": "3378548",
      "postDate": "12/18/2025 13:16:27",
      "content": "<p>I keep getting this error on submission:\nYour notebook generated a submission file with incorrect format. Some examples causing this are: wrong number of rows or columns, empty values, an incorrect data type for a value, or invalid submission values from what is expected. See more debugging tips.</p>\n<p>As far as I checked for the submission.csv it properly generates a submission in <code>/kaggle/working/submission.csv</code></p>\n<p>And it contains the correctly formatted content for the given (one) test image in test_images directory:</p>\n<pre><code>Submission shape: (1, 2)\nColumns: ['case_id', 'annotation']\n</code></pre>\n<p>Am I missing anything here? It keeps getting rejected with Submission scoring error. Would appreciate suggestions for this fix.</p>",
      "rawMarkdown": "I keep getting this error on submission:\nYour notebook generated a submission file with incorrect format. Some examples causing this are: wrong number of rows or columns, empty values, an incorrect data type for a value, or invalid submission values from what is expected. See more debugging tips.\n\nAs far as I checked for the submission.csv it properly generates a submission in `/kaggle/working/submission.csv`\n\nAnd it contains the correctly formatted content for the given (one) test image in test_images directory:\n```\nSubmission shape: (1, 2)\nColumns: ['case_id', 'annotation']\n```\n\nAm I missing anything here? It keeps getting rejected with Submission scoring error. Would appreciate suggestions for this fix.",
      "votes": null
    },
    {
      "id": "3378554",
      "postDate": "12/18/2025 13:29:19",
      "content": "<p>Try running it also on the train and supplemental images before submitting to verify it will work👍</p>",
      "rawMarkdown": "Try running it also on the train and supplemental images before submitting to verify it will work👍",
      "votes": null
    },
    {
      "id": "3378579",
      "postDate": "12/18/2025 14:46:20",
      "content": "<p>Thanks. Just to confirm, you mean I should run the inference on the images in the following directories in one go and generate a combined submission.csv with all these files: </p>\n<ul>\n<li>test_images</li>\n<li>train_images</li>\n<li>supplemental images</li>\n</ul>\n<p>?</p>",
      "rawMarkdown": "Thanks. Just to confirm, you mean I should run the inference on the images in the following directories in one go and generate a combined submission.csv with all these files: \n- test_images\n- train_images\n- supplemental images\n\n?",
      "votes": null
    },
    {
      "id": "3378961",
      "postDate": "12/18/2025 22:39:37",
      "content": "<p>Yea. When there is a scoring error, to test out if my submission works correctly, I run it on the supplemental images + 50 random authentic train + 50 random forged train. If there is an error with the submission, something usually will come out there.</p>\n<p>Perhaps the problem is maybe with how you format the annotation. Make sure it's RLE-Encoded correctly. Because I imagine your submission in the test is predicting 'authentic' so you might not see if there's an error there</p>",
      "rawMarkdown": "Yea. When there is a scoring error, to test out if my submission works correctly, I run it on the supplemental images + 50 random authentic train + 50 random forged train. If there is an error with the submission, something usually will come out there.\n\nPerhaps the problem is maybe with how you format the annotation. Make sure it's RLE-Encoded correctly. Because I imagine your submission in the test is predicting 'authentic' so you might not see if there's an error there",
      "votes": null
    },
    {
      "id": "3379231",
      "postDate": "12/19/2025 12:11:41",
      "content": "<p>Thanks for helping. I tried this approach as well but still not getting any specific error in the execution log. When I run it, the notebook runs successfully but on submission it returns submission scoring error.</p>\n<p>An example output of one of the submissions I tried:</p>\n<pre><code>Submission file saved to: /kaggle/working/submission.csv\nTotal rows: 1\n\nSubmission preview:\n  case_id annotation\n0      45  authentic\n\nStatistics:\n  Authentic images: 1\n  Forged images: 0\n  Forgery rate: 0.00%\n</code></pre>\n<p>Any thoughts?</p>",
      "rawMarkdown": "Thanks for helping. I tried this approach as well but still not getting any specific error in the execution log. When I run it, the notebook runs successfully but on submission it returns submission scoring error.\n\nAn example output of one of the submissions I tried:\n```\nSubmission file saved to: /kaggle/working/submission.csv\nTotal rows: 1\n\nSubmission preview:\n  case_id annotation\n0      45  authentic\n\nStatistics:\n  Authentic images: 1\n  Forged images: 0\n  Forgery rate: 0.00%\n```\n\nAny thoughts?",
      "votes": null
    },
    {
      "id": "3379243",
      "postDate": "12/19/2025 12:35:30",
      "content": "<p>Try putting your threshold way down so you actually predict an RLE string instead of authentic. Then instead of submitting, run the of1 score function yourself against the train’s ground truth to make sure the scoring runs correctly</p>",
      "rawMarkdown": "Try putting your threshold way down so you actually predict an RLE string instead of authentic. Then instead of submitting, run the of1 score function yourself against the train’s ground truth to make sure the scoring runs correctly",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3378554,
      "author_name": "returnofsputnik",
      "author_url": "",
      "post_date": "12/18/2025 13:29:19",
      "content": "<p>Try running it also on the train and supplemental images before submitting to verify it will work👍</p>",
      "votes": null,
      "replies": [
        {
          "id": 3378579,
          "author_name": "genesis96839",
          "author_url": "",
          "post_date": "12/18/2025 14:46:20",
          "content": "<p>Thanks. Just to confirm, you mean I should run the inference on the images in the following directories in one go and generate a combined submission.csv with all these files: </p>\n<ul>\n<li>test_images</li>\n<li>train_images</li>\n<li>supplemental images</li>\n</ul>\n<p>?</p>",
          "votes": null,
          "replies": [
            {
              "id": 3378961,
              "author_name": "returnofsputnik",
              "author_url": "",
              "post_date": "12/18/2025 22:39:37",
              "content": "<p>Yea. When there is a scoring error, to test out if my submission works correctly, I run it on the supplemental images + 50 random authentic train + 50 random forged train. If there is an error with the submission, something usually will come out there.</p>\n<p>Perhaps the problem is maybe with how you format the annotation. Make sure it's RLE-Encoded correctly. Because I imagine your submission in the test is predicting 'authentic' so you might not see if there's an error there</p>",
              "votes": null,
              "replies": [
                {
                  "id": 3379231,
                  "author_name": "genesis96839",
                  "author_url": "",
                  "post_date": "12/19/2025 12:11:41",
                  "content": "<p>Thanks for helping. I tried this approach as well but still not getting any specific error in the execution log. When I run it, the notebook runs successfully but on submission it returns submission scoring error.</p>\n<p>An example output of one of the submissions I tried:</p>\n<pre><code>Submission file saved to: /kaggle/working/submission.csv\nTotal rows: 1\n\nSubmission preview:\n  case_id annotation\n0      45  authentic\n\nStatistics:\n  Authentic images: 1\n  Forged images: 0\n  Forgery rate: 0.00%\n</code></pre>\n<p>Any thoughts?</p>",
                  "votes": null,
                  "replies": [
                    {
                      "id": 3379243,
                      "author_name": "returnofsputnik",
                      "author_url": "",
                      "post_date": "12/19/2025 12:35:30",
                      "content": "<p>Try putting your threshold way down so you actually predict an RLE string instead of authentic. Then instead of submitting, run the of1 score function yourself against the train’s ground truth to make sure the scoring runs correctly</p>",
                      "votes": null,
                      "replies": []
                    }
                  ]
                }
              ]
            }
          ]
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "3378548": "I keep getting this error on submission:\nYour notebook generated a submission file with incorrect format. Some examples causing this are: wrong number of rows or columns, empty values, an incorrect data type for a value, or invalid submission values from what is expected. See more debugging tips.\n\nAs far as I checked for the submission.csv it properly generates a submission in `/kaggle/working/submission.csv`\n\nAnd it contains the correctly formatted content for the given (one) test image in test_images directory:\n```\nSubmission shape: (1, 2)\nColumns: ['case_id', 'annotation']\n```\n\nAm I missing anything here? It keeps getting rejected with Submission scoring error. Would appreciate suggestions for this fix.",
    "3378554": "Try running it also on the train and supplemental images before submitting to verify it will work👍",
    "3378579": "Thanks. Just to confirm, you mean I should run the inference on the images in the following directories in one go and generate a combined submission.csv with all these files: \n- test_images\n- train_images\n- supplemental images\n\n?",
    "3378961": "Yea. When there is a scoring error, to test out if my submission works correctly, I run it on the supplemental images + 50 random authentic train + 50 random forged train. If there is an error with the submission, something usually will come out there.\n\nPerhaps the problem is maybe with how you format the annotation. Make sure it's RLE-Encoded correctly. Because I imagine your submission in the test is predicting 'authentic' so you might not see if there's an error there",
    "3379231": "Thanks for helping. I tried this approach as well but still not getting any specific error in the execution log. When I run it, the notebook runs successfully but on submission it returns submission scoring error.\n\nAn example output of one of the submissions I tried:\n```\nSubmission file saved to: /kaggle/working/submission.csv\nTotal rows: 1\n\nSubmission preview:\n  case_id annotation\n0      45  authentic\n\nStatistics:\n  Authentic images: 1\n  Forged images: 0\n  Forgery rate: 0.00%\n```\n\nAny thoughts?",
    "3379243": "Try putting your threshold way down so you actually predict an RLE string instead of authentic. Then instead of submitting, run the of1 score function yourself against the train’s ground truth to make sure the scoring runs correctly"
  },
  "source": "meta"
}