{
  "id": 613246,
  "title": "CRITICAL: Submission Scoring Error (Wrong Number of Rows) - Sample Submission Parsing Issue",
  "url": "/competitions/recodai-luc-scientific-image-forgery-detection/discussion/613246",
  "author_name": "",
  "post_date": "2025-10-25T08:56:38.826901200Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>I am encountering a persistent Submission Scoring Error due to an incorrect row count, which appears to be caused by the environment failing to read the sample_submission.csv file correctly.</p>\n<p>The Expected Issue:<br>\nThe competition requires a submission file with approximately 1,100 rows (one for each public test image ID).</p>\n<p>The Observed Failure (Guaranteed Reproducible):<br>\nWhen running the notebook (even the official fail-safe, model-free submission method that only reads the CSV and defaults annotations to 'authentic'), pandas.read_csv is only parsing 1 data row from sample_submission.csv.</p>\n<p>Debugging Steps Performed:</p>\n<p>Code Used: Simple pd.read_csv(SAMPLE_SUBMISSION_PATH) followed by saving the resulting DataFrame.</p>\n<p>Output Log:</p>\n<p>--- Generating Guaranteed Submission Row Count (Authentic Only) ---<br>\nFound 1 required image (ID) for submission.<br>\n✅ SUCCESS: Submission file 'submission.csv' created with one prediction.</p>\n<p>Result: The resulting submission.csv has a single data row, causing the scoring engine to immediately reject it with the error: \"wrong number of rows or columns.\"</p>\n<p>Request for Assistance:</p>\n<p>This failure to load the full lis list of required IDs from the official sample_submission.csv is preventing all submissions. Could the competition hosts please investigate the integrity or encoding of the sample_submission.csv file within the code execution environment? Alternatively, please provide an official list of all public test set IDs.</p>\n<p>Thank you for your prompt attention to this matter.</p>",
  "messages": [
    {
      "id": "3306786",
      "postDate": "10/25/2025 08:56:38",
      "content": "<p>I am encountering a persistent Submission Scoring Error due to an incorrect row count, which appears to be caused by the environment failing to read the sample_submission.csv file correctly.</p>\n<p>The Expected Issue:<br>\nThe competition requires a submission file with approximately 1,100 rows (one for each public test image ID).</p>\n<p>The Observed Failure (Guaranteed Reproducible):<br>\nWhen running the notebook (even the official fail-safe, model-free submission method that only reads the CSV and defaults annotations to 'authentic'), pandas.read_csv is only parsing 1 data row from sample_submission.csv.</p>\n<p>Debugging Steps Performed:</p>\n<p>Code Used: Simple pd.read_csv(SAMPLE_SUBMISSION_PATH) followed by saving the resulting DataFrame.</p>\n<p>Output Log:</p>\n<p>--- Generating Guaranteed Submission Row Count (Authentic Only) ---<br>\nFound 1 required image (ID) for submission.<br>\n✅ SUCCESS: Submission file 'submission.csv' created with one prediction.</p>\n<p>Result: The resulting submission.csv has a single data row, causing the scoring engine to immediately reject it with the error: \"wrong number of rows or columns.\"</p>\n<p>Request for Assistance:</p>\n<p>This failure to load the full lis list of required IDs from the official sample_submission.csv is preventing all submissions. Could the competition hosts please investigate the integrity or encoding of the sample_submission.csv file within the code execution environment? Alternatively, please provide an official list of all public test set IDs.</p>\n<p>Thank you for your prompt attention to this matter.</p>",
      "rawMarkdown": "I am encountering a persistent Submission Scoring Error due to an incorrect row count, which appears to be caused by the environment failing to read the sample_submission.csv file correctly.\n\nThe Expected Issue:\nThe competition requires a submission file with approximately 1,100 rows (one for each public test image ID).\n\nThe Observed Failure (Guaranteed Reproducible):\nWhen running the notebook (even the official fail-safe, model-free submission method that only reads the CSV and defaults annotations to 'authentic'), pandas.read_csv is only parsing 1 data row from sample_submission.csv.\n\nDebugging Steps Performed:\n\nCode Used: Simple pd.read_csv(SAMPLE_SUBMISSION_PATH) followed by saving the resulting DataFrame.\n\nOutput Log:\n\n--- Generating Guaranteed Submission Row Count (Authentic Only) ---\nFound 1 required image (ID) for submission.\n✅ SUCCESS: Submission file 'submission.csv' created with one prediction.\n\n\nResult: The resulting submission.csv has a single data row, causing the scoring engine to immediately reject it with the error: \"wrong number of rows or columns.\"\n\nRequest for Assistance:\n\nThis failure to load the full lis list of required IDs from the official sample_submission.csv is preventing all submissions. Could the competition hosts please investigate the integrity or encoding of the sample_submission.csv file within the code execution environment? Alternatively, please provide an official list of all public test set IDs.\n\nThank you for your prompt attention to this matter.",
      "votes": null
    },
    {
      "id": "3375758",
      "postDate": "12/13/2025 04:10:01",
      "content": "<p>while creating submission , do predtion  on  test images folder </p>",
      "rawMarkdown": "while creating submission , do predtion  on  test images folder",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3375758,
      "author_name": "ankurksingh1312",
      "author_url": "",
      "post_date": "12/13/2025 04:10:01",
      "content": "<p>while creating submission , do predtion  on  test images folder </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3306786": "I am encountering a persistent Submission Scoring Error due to an incorrect row count, which appears to be caused by the environment failing to read the sample_submission.csv file correctly.\n\nThe Expected Issue:\nThe competition requires a submission file with approximately 1,100 rows (one for each public test image ID).\n\nThe Observed Failure (Guaranteed Reproducible):\nWhen running the notebook (even the official fail-safe, model-free submission method that only reads the CSV and defaults annotations to 'authentic'), pandas.read_csv is only parsing 1 data row from sample_submission.csv.\n\nDebugging Steps Performed:\n\nCode Used: Simple pd.read_csv(SAMPLE_SUBMISSION_PATH) followed by saving the resulting DataFrame.\n\nOutput Log:\n\n--- Generating Guaranteed Submission Row Count (Authentic Only) ---\nFound 1 required image (ID) for submission.\n✅ SUCCESS: Submission file 'submission.csv' created with one prediction.\n\n\nResult: The resulting submission.csv has a single data row, causing the scoring engine to immediately reject it with the error: \"wrong number of rows or columns.\"\n\nRequest for Assistance:\n\nThis failure to load the full lis list of required IDs from the official sample_submission.csv is preventing all submissions. Could the competition hosts please investigate the integrity or encoding of the sample_submission.csv file within the code execution environment? Alternatively, please provide an official list of all public test set IDs.\n\nThank you for your prompt attention to this matter.",
    "3375758": "while creating submission , do predtion  on  test images folder"
  },
  "source": "meta"
}