{
  "id": 388415,
  "title": "Previous successful submission notebook now throwing Memory Error",
  "url": "/competitions/rsna-breast-cancer-detection/discussion/388415",
  "author_name": "sandy1112",
  "post_date": "2023-02-17T10:36:59.285000",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I am facing a strange issue where a previously successful submission when submitted again with only a single change of updating the probability threshold value keeps throwing Out of memory error. Had about 7 such issues in the last two days. If a sub was successful few days back, a simple change of threshold value should should not cause this issue. Anyone else observed this? Wonder if there has been some update in the hidden test images or if some other issue which others could guess? 🤔</p>",
  "messages": [
    {
      "id": 2149720,
      "postDate": "2023-02-18T15:48:49.350Z",
      "content": "<p>my guess is that  either the kernel hardware not same or the score too good then using oom to cheat u.</p>\n<p>I have the same problem.</p>",
      "rawMarkdown": "my guess is that  either the kernel hardware not same or the score too good then using oom to cheat u.\n\nI have the same problem.",
      "votes": 1,
      "replies": [
        {
          "id": 2149724,
          "postDate": "2023-02-18T15:56:39.073Z",
          "content": "<p>my suspicion is also same as you…kernel hardware allocated may not be of the same nature everytime a new notebook instance is spawned. I have seen great amount of variance in the way parallel processing works between multiple runs of the same notebook (without a single change). </p>",
          "rawMarkdown": "my suspicion is also same as you...kernel hardware allocated may not be of the same nature everytime a new notebook instance is spawned. I have seen great amount of variance in the way parallel processing works between multiple runs of the same notebook (without a single change). "
        }
      ]
    },
    {
      "id": 2148302,
      "postDate": "2023-02-17T10:36:59.287Z",
      "content": "<p>I am facing a strange issue where a previously successful submission when submitted again with only a single change of updating the probability threshold value keeps throwing Out of memory error. Had about 7 such issues in the last two days. If a sub was successful few days back, a simple change of threshold value should should not cause this issue. Anyone else observed this? Wonder if there has been some update in the hidden test images or if some other issue which others could guess? 🤔</p>",
      "rawMarkdown": "I am facing a strange issue where a previously successful submission when submitted again with only a single change of updating the probability threshold value keeps throwing Out of memory error. Had about 7 such issues in the last two days. If a sub was successful few days back, a simple change of threshold value should should not cause this issue. Anyone else observed this? Wonder if there has been some update in the hidden test images or if some other issue which others could guess? 🤔\n\n",
      "votes": 1
    }
  ],
  "comments": [
    {
      "id": 2149720,
      "author_name": "dragon zhang",
      "author_url": "",
      "post_date": "2023-02-18T15:48:49.350000",
      "content": "<p>my guess is that  either the kernel hardware not same or the score too good then using oom to cheat u.</p>\n<p>I have the same problem.</p>",
      "votes": 1,
      "replies": [
        {
          "id": 2149724,
          "author_name": "sandy1112",
          "author_url": "",
          "post_date": "2023-02-18T15:56:39.073000",
          "content": "<p>my suspicion is also same as you…kernel hardware allocated may not be of the same nature everytime a new notebook instance is spawned. I have seen great amount of variance in the way parallel processing works between multiple runs of the same notebook (without a single change). </p>",
          "votes": 0,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2149720": "my guess is that  either the kernel hardware not same or the score too good then using oom to cheat u.\n\nI have the same problem.",
    "2148302": "I am facing a strange issue where a previously successful submission when submitted again with only a single change of updating the probability threshold value keeps throwing Out of memory error. Had about 7 such issues in the last two days. If a sub was successful few days back, a simple change of threshold value should should not cause this issue. Anyone else observed this? Wonder if there has been some update in the hidden test images or if some other issue which others could guess? 🤔\n\n"
  }
}