{
  "id": 478270,
  "title": "This is too much!!!!",
  "url": "/competitions/home-credit-credit-risk-model-stability/discussion/478270",
  "author_name": "",
  "post_date": "2024-02-19T23:21:16.707889Z",
  "votes": -4,
  "comment_count": 2,
  "views": 0,
  "content": "<p>every time we solve an error another one pop up what does notebook threw exception mean exactly?? how am i supposed to know what the problem is when the data is hidden and the errors are unclear??? it ran correctly its not a memory issue before anyone say it is as always</p>",
  "messages": [
    {
      "id": "2659524",
      "postDate": "02/19/2024 23:21:16",
      "content": "<p>every time we solve an error another one pop up what does notebook threw exception mean exactly?? how am i supposed to know what the problem is when the data is hidden and the errors are unclear??? it ran correctly its not a memory issue before anyone say it is as always</p>",
      "rawMarkdown": "every time we solve an error another one pop up what does notebook threw exception mean exactly?? how am i supposed to know what the problem is when the data is hidden and the errors are unclear??? it ran correctly its not a memory issue before anyone say it is as always",
      "votes": null
    },
    {
      "id": "2659689",
      "postDate": "02/20/2024 04:53:39",
      "content": "<p>When you run a notebook, it runs on just a tiny sample of test data (10 examples), when you submit the notebook it is evaluated on the full public dataset, so any error is likely due to that difference. Debugging is more difficult, I would try to reduce the solution to a simple and stable version and then incrementally add the more tricky parts.</p>",
      "rawMarkdown": "When you run a notebook, it runs on just a tiny sample of test data (10 examples), when you submit the notebook it is evaluated on the full public dataset, so any error is likely due to that difference. Debugging is more difficult, I would try to reduce the solution to a simple and stable version and then incrementally add the more tricky parts.",
      "votes": null
    },
    {
      "id": "2660406",
      "postDate": "02/20/2024 15:29:01",
      "content": "<p>Try to run your inference code with train dataset and see if you get any OOM errors (monitor your RAM constantly). </p>\n<p>From my experience 95% of Notebook Threw Exception errors are due to OOM</p>",
      "rawMarkdown": "Try to run your inference code with train dataset and see if you get any OOM errors (monitor your RAM constantly). \n\nFrom my experience 95% of Notebook Threw Exception errors are due to OOM",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2659689,
      "author_name": "lohmaa",
      "author_url": "",
      "post_date": "02/20/2024 04:53:39",
      "content": "<p>When you run a notebook, it runs on just a tiny sample of test data (10 examples), when you submit the notebook it is evaluated on the full public dataset, so any error is likely due to that difference. Debugging is more difficult, I would try to reduce the solution to a simple and stable version and then incrementally add the more tricky parts.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 2660406,
      "author_name": "imeintanis",
      "author_url": "",
      "post_date": "02/20/2024 15:29:01",
      "content": "<p>Try to run your inference code with train dataset and see if you get any OOM errors (monitor your RAM constantly). </p>\n<p>From my experience 95% of Notebook Threw Exception errors are due to OOM</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2659524": "every time we solve an error another one pop up what does notebook threw exception mean exactly?? how am i supposed to know what the problem is when the data is hidden and the errors are unclear??? it ran correctly its not a memory issue before anyone say it is as always",
    "2659689": "When you run a notebook, it runs on just a tiny sample of test data (10 examples), when you submit the notebook it is evaluated on the full public dataset, so any error is likely due to that difference. Debugging is more difficult, I would try to reduce the solution to a simple and stable version and then incrementally add the more tricky parts.",
    "2660406": "Try to run your inference code with train dataset and see if you get any OOM errors (monitor your RAM constantly). \n\nFrom my experience 95% of Notebook Threw Exception errors are due to OOM"
  },
  "source": "meta"
}