{
  "id": 589410,
  "title": "Submission Error: 'Notebook Threw Exception' – What Could Be Wrong",
  "url": "/competitions/cmi-detect-behavior-with-sensor-data/discussion/589410",
  "author_name": "",
  "post_date": "2025-07-12T16:26:40.099028800Z",
  "votes": 1,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Hey everyone,<br>\nI followed some open-source notebooks and wrote my inference code (a predict function), and also added the CMI inference server code snippet. However, when I submit the final prediction to the leaderboard, I get an error message saying: \"Notebook Threw Exception.\"<br>\nHow can I find out what went wrong? Am I missing something?</p>",
  "messages": [
    {
      "id": "3247390",
      "postDate": "07/12/2025 16:26:40",
      "content": "<p>Hey everyone,<br>\nI followed some open-source notebooks and wrote my inference code (a predict function), and also added the CMI inference server code snippet. However, when I submit the final prediction to the leaderboard, I get an error message saying: \"Notebook Threw Exception.\"<br>\nHow can I find out what went wrong? Am I missing something?</p>",
      "rawMarkdown": "Hey everyone,\nI followed some open-source notebooks and wrote my inference code (a predict function), and also added the CMI inference server code snippet. However, when I submit the final prediction to the leaderboard, I get an error message saying: \"Notebook Threw Exception.\"\nHow can I find out what went wrong? Am I missing something?",
      "votes": null
    },
    {
      "id": "3247400",
      "postDate": "07/12/2025 16:35:00",
      "content": "<p>Try testing your predict function with the samples in train.csv one by one, usually you'll find the reason</p>",
      "rawMarkdown": "Try testing your predict function with the samples in train.csv one by one, usually you'll find the reason",
      "votes": null
    },
    {
      "id": "3247478",
      "postDate": "07/12/2025 19:53:28",
      "content": "<p>because the test set in your notebook have only 2 sequences and both have thm/tof data</p>\n<p>while half of the  private test set have imu features only and consider the other values null, so your code probably have problems with the imu ony data</p>\n<p>to debug it for one of the 2 sequences nullify all non-imu features, it should work and show you the error when you run it</p>",
      "rawMarkdown": "because the test set in your notebook have only 2 sequences and both have thm/tof data\n\nwhile half of the  private test set have imu features only and consider the other values null, so your code probably have problems with the imu ony data\n\nto debug it for one of the 2 sequences nullify all non-imu features, it should work and show you the error when you run it",
      "votes": null
    },
    {
      "id": "3247572",
      "postDate": "07/13/2025 02:37:04",
      "content": "<p>try clear output</p>",
      "rawMarkdown": "try clear output",
      "votes": null
    },
    {
      "id": "3247676",
      "postDate": "07/13/2025 07:46:36",
      "content": "<ol>\n<li>inference predict function </li>\n<li>computational time </li>\n<li>notebook internet </li>\n<li>wait for the notebook to save</li>\n<li>Try again</li>\n</ol>",
      "rawMarkdown": "1. inference predict function \n2. computational time \n3. notebook internet \n4. wait for the notebook to save\n5. Try again",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3247400,
      "author_name": "minerppdy",
      "author_url": "",
      "post_date": "07/12/2025 16:35:00",
      "content": "<p>Try testing your predict function with the samples in train.csv one by one, usually you'll find the reason</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3247478,
      "author_name": "tahaalshatiri",
      "author_url": "",
      "post_date": "07/12/2025 19:53:28",
      "content": "<p>because the test set in your notebook have only 2 sequences and both have thm/tof data</p>\n<p>while half of the  private test set have imu features only and consider the other values null, so your code probably have problems with the imu ony data</p>\n<p>to debug it for one of the 2 sequences nullify all non-imu features, it should work and show you the error when you run it</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3247572,
      "author_name": "hiroarimatsumoto",
      "author_url": "",
      "post_date": "07/13/2025 02:37:04",
      "content": "<p>try clear output</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3247676,
      "author_name": "sivaprasadnaidu",
      "author_url": "",
      "post_date": "07/13/2025 07:46:36",
      "content": "<ol>\n<li>inference predict function </li>\n<li>computational time </li>\n<li>notebook internet </li>\n<li>wait for the notebook to save</li>\n<li>Try again</li>\n</ol>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3247390": "Hey everyone,\nI followed some open-source notebooks and wrote my inference code (a predict function), and also added the CMI inference server code snippet. However, when I submit the final prediction to the leaderboard, I get an error message saying: \"Notebook Threw Exception.\"\nHow can I find out what went wrong? Am I missing something?",
    "3247400": "Try testing your predict function with the samples in train.csv one by one, usually you'll find the reason",
    "3247478": "because the test set in your notebook have only 2 sequences and both have thm/tof data\n\nwhile half of the  private test set have imu features only and consider the other values null, so your code probably have problems with the imu ony data\n\nto debug it for one of the 2 sequences nullify all non-imu features, it should work and show you the error when you run it",
    "3247572": "try clear output",
    "3247676": "1. inference predict function \n2. computational time \n3. notebook internet \n4. wait for the notebook to save\n5. Try again"
  },
  "source": "meta"
}